array:20 [
  "pii" => "13051580"
  "issn" => "02139111"
  "estado" => "S300"
  "fechaPublicacion" => "2003-10-01"
  "documento" => "article"
  "crossmark" => 0
  "licencia" => "http://www.elsevier.com/open-access/userlicense/1.0/"
  "subdocumento" => "fla"
  "cita" => "Gac Sanit. 2003;17 Supl 2:85-90"
  "abierto" => array:3 [
    "ES" => true
    "ES2" => true
    "LATM" => true
  ]
  "gratuito" => true
  "lecturas" => array:2 [
    "total" => 4035
    "formatos" => array:3 [
      "EPUB" => 161
      "HTML" => 3460
      "PDF" => 414
    ]
  ]
  "itemSiguiente" => array:16 [
    "pii" => "13051581"
    "issn" => "02139111"
    "estado" => "S300"
    "fechaPublicacion" => "2003-10-01"
    "documento" => "article"
    "crossmark" => 0
    "licencia" => "http://www.elsevier.com/open-access/userlicense/1.0/"
    "subdocumento" => "fla"
    "cita" => "Gac Sanit. 2003;17 Supl 2:90-100"
    "abierto" => array:3 [
      "ES" => true
      "ES2" => true
      "LATM" => true
    ]
    "gratuito" => true
    "lecturas" => array:2 [
      "total" => 4270
      "formatos" => array:3 [
        "EPUB" => 148
        "HTML" => 3479
        "PDF" => 643
      ]
    ]
    "en" => array:8 [
      "idiomaDefecto" => true
      "titulo" => "Pòsters : Vigilancia epidemiológica"
      "tienePdf" => "en"
      "tieneTextoCompleto" => "en"
      "paginas" => array:1 [
        0 => array:2 [
          "paginaInicial" => "90"
          "paginaFinal" => "100"
        ]
      ]
      "titulosAlternativos" => array:1 [
        "en" => array:1 [
          "titulo" => "Surveillance"
        ]
      ]
      "contieneTextoCompleto" => array:1 [
        "en" => true
      ]
      "contienePdf" => array:1 [
        "en" => true
      ]
    ]
    "idiomaDefecto" => "en"
    "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/13051581?idApp=WGSE"
    "url" => "/02139111/00000017000000S2/v0_201302051404/13051581/v0_201302051405/en/main.assets"
  ]
  "itemAnterior" => array:16 [
    "pii" => "13051579"
    "issn" => "02139111"
    "estado" => "S300"
    "fechaPublicacion" => "2003-10-01"
    "documento" => "article"
    "crossmark" => 0
    "licencia" => "http://www.elsevier.com/open-access/userlicense/1.0/"
    "subdocumento" => "fla"
    "cita" => "Gac Sanit. 2003;17 Supl 2:81-4"
    "abierto" => array:3 [
      "ES" => true
      "ES2" => true
      "LATM" => true
    ]
    "gratuito" => true
    "lecturas" => array:2 [
      "total" => 2642
      "formatos" => array:3 [
        "EPUB" => 141
        "HTML" => 2064
        "PDF" => 437
      ]
    ]
    "en" => array:8 [
      "idiomaDefecto" => true
      "titulo" => "Pòsters : Salud Infantil, salud reproductiva, género y salud"
      "tienePdf" => "en"
      "tieneTextoCompleto" => "en"
      "paginas" => array:1 [
        0 => array:2 [
          "paginaInicial" => "81"
          "paginaFinal" => "84"
        ]
      ]
      "titulosAlternativos" => array:1 [
        "en" => array:1 [
          "titulo" => "Child health, reproductive health, gender and health"
        ]
      ]
      "contieneTextoCompleto" => array:1 [
        "en" => true
      ]
      "contienePdf" => array:1 [
        "en" => true
      ]
    ]
    "idiomaDefecto" => "en"
    "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/13051579?idApp=WGSE"
    "url" => "/02139111/00000017000000S2/v0_201302051404/13051579/v0_201302051405/en/main.assets"
  ]
  "en" => array:9 [
    "idiomaDefecto" => true
    "titulo" => "Pòsters : Cáncer"
    "tieneTextoCompleto" => true
    "paginas" => array:1 [
      0 => array:2 [
        "paginaInicial" => "85"
        "paginaFinal" => "90"
      ]
    ]
    "titulosAlternativos" => array:1 [
      "en" => array:1 [
        "titulo" => "Cancer"
      ]
    ]
    "textoCompleto" => "<p class="elsevierStylePara"> Jueves 2 de Octubre &#47; Thursday 2&#44; October<br></br> 17&#58;00&#58;00 a&#47;to 18&#58;00&#58;00</p><p class="elsevierStylePara"><span class="elsevierStyleBold">111</span><span class="elsevierStyleBold">FACTORES QUE DETERMINAN EL USO DEL PAPANICOLAOU EN MUJERES MEXICANAS</span></p><p class="elsevierStylePara"> Rosa Mar&#237;a Ortiz Espinosa&#42;&#44; Sergio Mu&#241;oz Ju&#225;rez&#42;&#44; Socorro M&#225;rquez Maldonado&#42;&#42;&#44; Maria de los Angeles Moron Arella&#241;o&#42;&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Coordinaci&#243;n de Investigaci&#243;n&#44; Secretaria de Salud de Hidalgo&#44; Pachuca&#44; M&#233;xico&#46; &#42;&#42;Direcci&#243;n de Regulaci&#243;n Sanitaria&#44; Secretaria de Salud&#44; Pachuca&#44; M&#233;xico&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Antecedentes</span>&#58; El c&#225;ncer cervicouterino es la neoplasia m&#225;s frecuente en mujeres mexicanas y en Latinoam&#233;rica&#46; En M&#233;xico existe un programa poblacional desde hace 20 a&#241;os&#44; no obstante la mortalidad por esta causa continua siendo constante&#46; Existen factores relacionados con la cobertura y accesibilidad del servicio que influyen en la detecci&#243;n y tratamiento oportuno&#44; pero adem&#225;s se encuentran los relacionados con la aceptabilidad&#44; la tendencia e intensidad de uso del Papanicolaou &#40;PAP&#41; por parte de la poblaci&#243;n&#44; que influyen en la decisi&#243;n de no hacerse el PAP&#46; En el &#225;rea rural&#44; las posibilidades que tiene la mujer para decidir sobre su vida sexual y su auto cuidado son&#44; con frecuencia limitadas&#44; la capacidad general para negociar con &#233;xito las necesidades en torno a la salud de su propio organismo son limitadas&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Objetivo&#58;</span> Identificar los factores que determinan el uso del PAP&#46; Material y m&#233;todos&#58; Se realiz&#243; un dise&#241;o transversal anal&#237;tico y comparativo&#46; Se aplic&#243; una entrevista estructurada a usuarios de 15 a 49 a&#241;os&#44; de las unidades de primer nivel&#44; seleccionados aleatoriamente&#46; Se utiliz&#243; estad&#237;stica descriptiva&#44; Ji cuadrada y Regresi&#243;n log&#237;stica no condicional&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> Se encontr&#243; que el 31&#44;22&#37; de los encuestados desconoc&#237;an el PAP y su funci&#243;n&#44; de este grupo&#44; el 82&#46;5&#37; nunca se la hab&#237;an hecho&#46; La mayor proporci&#243;n de desconocimiento se observo en los hombres en comparaci&#243;n con las mujeres &#40;45&#37; vs 15&#44;49&#37;&#41;&#44; el desconocimiento en los usuarios analfabetos fue 33&#44;1&#37; y de 44&#44;3&#37; en los residentes de municipios de mayor marginaci&#243;n&#46; Cuatro de cada diez varones lo desconocen&#44; y en las mujeres dos de cada diez &#40;P &#61; 0&#44;004&#41;&#46; El 46&#37; de los usuarios ignoran los factores de riesgo asociados a cacu&#46; En las usuarias que refirieron nunca haberse efectuado un PAP&#44; las principales causas fueron&#44; porque no lo consideran necesario Raz&#243;n de Momios &#40;RM&#41; crudo de 2&#44;5 con intervalos de confianza &#40;IC&#41; de 95&#37;&#58; 1&#44;3 a 4&#46;8&#44; el personal no me da confianza RM 4&#44;1 IC 95&#37; 1&#44;3 a 12&#44;3&#46; El conocimiento del c&#225;ncer cervico uterino &#40;CaCu&#41; y el saber que es curable fue diferente&#44; en las alguna vez usuarias del Pap y aquellas nunca usuarias&#40;P &#61; 0&#46;000&#41;&#44; as&#237; como el conocimiento de la prevenci&#243;n del cacu &#40;P &#61; 0&#46;00&#41;&#46; Las variables asociadas a la demanda del PAP fueron el desconocimiento del CaCu Raz&#243;n de Momios &#40;RM&#41; 3&#44;6 IC al 95&#37; &#40;1&#46;7 a 7&#46;7&#41;&#44; desconocer que se puede evitar RM 1&#44;71 IC &#40;1&#46;19&#44;2&#46;4&#41;&#44; desconocimiento que es curable RM 3&#44;36 IC &#40;1&#46;8&#44;6&#46;2&#41; y ser analfabeta RM 4&#46;24 IC &#40;2&#46;1&#44;8&#46;3&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> El conocimiento previo de la neoplasia&#44; y el saber que es curable son factores primordiales que se deben de incluir en las acciones de informaci&#243;n&#44; educaci&#243;n y comunicaci&#243;n &#40;IEC&#41; de los programas de prevenci&#243;n y destacar la necesidad de dise&#241;ar estrategias donde se difunda los ben&#233;ficos del PAP&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">112</span><span class="elsevierStyleBold">SMOKING&#44; ALCOHOL&#44; AND DENTITION IN THE EPIDEMIOLOGY OF ORAL CANCER IN POLAND</span></p><p class="elsevierStylePara"> Jolanta Lissowska&#42;&#44; Agnieszka Pilarska&#42;&#42;&#44; Pawel Pilarski&#42;&#42;&#44; Danuta Samolczyk-Wanyura&#42;&#42;&#44; Janusz Piekarczyk&#42;&#42;&#44; Alicja Bardin-Mikolajczak&#42;&#44; Witold Zatonski&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Dept&#46; of Cancer Epidemiology and Prevention&#44; Cancer Center &#38; M Sklodowska-Curie Institute of Oncology&#44; Warsaw&#44; Poland&#46; &#42;&#42;2 2nd Maxillofacial Surgery Clinic&#44; Medical Academy&#44; Warsaw&#44; Poland&#46;</span></p><p class="elsevierStylePara"> The role of smoking&#44; drinking&#44; and dental care on the risk of oral and pharyngeal cancer was investigated in a case-control study conducted in Warsaw&#44; Poland&#46; Cases were 122 patients &#40;including 44 females&#41; aged 23-80 years with incident&#44; histologically confirmed cancer of oral cavity and pharynx&#46; Controls were 124 subjects &#40;including 52 females&#41; admitted to the hospital for different non-neoplastic conditions unrelated to tobacco and alcohol consumption&#44; frequency matched to cases by age and sex&#46; Smoking and drinking were strongly associated with an increased risk of oral cancer&#46; Among consumers of both products&#44; risks of oral cancer tended to combine in a multiplicative fashion and were increased more than 14-fold among those who consumed more than 15 cigarettes and 7 or more drinks per day&#46; Cessation of smoking was associated with reduced risk of this cancer&#46; The risks varied by type of cigarettes smoked&#44; being lower among those consuming filtered cigarettes only &#40;OR&#61;1&#46;6&#41; than non-filter &#40;OR&#61;6&#46;5&#41; or mixed &#40;OR&#61;4&#46;2&#41; cigarettes&#46; After adjustment for tobacco smoking and alcohol drinking&#44; poor dentition as reflected by missing teeth&#44; frequency of dental check-ups and frequency of teeth brushing emerged as a strong risk factors&#46; Number of missing teeth and frequency of dental check-ups and frequency of tooth brushing showed increased ORs of 9&#46;8&#44; 11&#46;9 and 3&#46;2 respectively&#46; Denture wearing per se did not affect oral cancer risk&#46; In terms of attributable risk&#44; smoking accounted for 57&#37; of oral cancer cases in Poland&#44; alcohol for 31&#37;&#46; Attributable risks for low frequency of tooth brushing and dental check-ups were 56&#37; and 47&#37; respectively&#46; In conclusion&#44; smoking and drinking cessation are likely to be effective preventive measures against oral cancer&#46; These findings indicate also that poor oral hygiene may be independent risk factor&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">113</span><span class="elsevierStyleBold">DECLINING MORTALITY RATES FOR NONMELANOMA SKIN CANCERS IN WEST GERMANY&#44; 1968 THROUGH 1999&#46; AN ANALYSIS OF 11&#46;226 NONMELANOMA SKIN CANCER DEATHS</span></p><p class="elsevierStylePara"> Andreas Stang&#44; Karl-Heinz J&#246;ckel</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Epidemiology Unit&#44; Medical Faculaty&#44; University of Essen&#44; Essen&#44; Germany&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Purpose&#58;</span> Since the primary source of data for cancer registries is the inpatient hospital file&#44; routinely collected statistics on nonmelanoma skin cancer &#40;NMSC&#41; are usually incomplete and not comparable with other forms of cancer&#46; We therefore examined time trends of the nonmelanoma skin cancer mortality for the territory of West-Germany including a population of about 66 million people&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> We analysed the nonmelanocytic skin cancer mortality data &#40;1968-1999&#41; from West-Germany including West-Berlin&#46; We calculated age-specific and age-standardized mortality rates &#40;World Standard Population&#41; and used Poisson regression to estimate underlying age&#44; cohort and period effect&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> From 1968 &#40;population size of the territory of the Federal State of Germany before the reunification&#58; 60&#46;0 million&#41; through 1999 &#40;population size&#58; 66&#46;9 million&#41;&#44; about 11&#46;226 deaths were attributed to NMSC&#46; The NMSC mortality was greater among men than among women throughout the period studied&#46; The estimated percent annual decrease of the age-standardized nonmelanocytic skin cancer mortality rate was -2&#46;3&#37; &#40;95&#37;CI -2&#46;6&#59; -1&#46;9&#41; among men and -3&#46;5&#37; &#40;95&#37;CI&#58; -4&#46;0&#59; -3&#46;1&#41; among women during the period 1968 through 1999&#46; This decline is mainly due to a rate decrease in people aged 80 years or more&#46; The age-specific estimated annual percent changes of the NMSC mortality rates indicate that the greater decline among women is mainly due to a greater rate decrease in women aged 80 years or more&#46; The change in nonmelanoma skin cancer mortality rates was best explained by age-&#44; cohort- and period effects&#46; The age-specific proportions of skin cancer deaths attributed to NMSC declined in people aged 50 year or more from 1968 through 1999&#46; In the early period from 1968 through 1979&#44; about 58&#37; of the male skin cancer deaths and 61&#37; of the female skin cancer deaths in people age 80 years or more were attributed to NSMC&#46; These proportions declined to 33&#37; and 29&#37;&#44; respectively in the latest period from 1990 through 1999&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> The nonmelanoma skin cancer mortality in West-Germany showed a continuous decrease from 1968 through 1999&#46; The favourable mortality decline by birth cohort in the most recent birth cohort is an important indicator of a likely decline in mortality over the next years&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">114</span><span class="elsevierStyleBold">POLYMORPHISMS G691S &#47; S904S OF RET AS GENETIC MODIFIERS IN CANCER PATIENTS FROM FAMILIES WITH MULTIPLE ENDOCRINE NEOPLASIA TYPE 2A</span></p><p class="elsevierStylePara"> Marina Poll&#225;n<span class="elsevierStyleSup">1</span>&#44; Mercedes Robledo<span class="elsevierStyleSup">2</span>&#44; Laura Gil<span class="elsevierStyleSup">3</span>&#44; Arancha Cebri&#225;n<span class="elsevierStyleSup">2</span>&#44; Sergio Ruiz<span class="elsevierStyleSup">2</span>&#44; Marta Aza&#241;edo<span class="elsevierStyleSup">2</span>&#44; Javier Ben&#237;tez<span class="elsevierStyleSup">2</span>&#44; Javier Men&#225;rguez<span class="elsevierStyleSup">4</span>&#44; Jose Mar&#237;a Rojas<span class="elsevierStyleSup">3</span></p><p class="elsevierStylePara"><span class="elsevierStyleItalic"><span class="elsevierStyleSup">1</span>&#193;rea de Epidemiolog&#237;a Ambiental y C&#225;ncer&#44; Centro Nacional de Epidemiolog&#237;a del ISCIII&#46;&#44; Madrid&#44; Spain&#46; <span class="elsevierStyleSup">2</span>Unidad de Gen&#233;tica&#44; Centro Nacional de Investigaciones Oncol&#243;gicas &#40;CNIO&#41;&#44; Madrid&#44; Spain&#46; <span class="elsevierStyleSup">3</span>Unidad de Biolog&#237;a Celular&#44; Centro Nacional de Microbiolog&#237;a del ISCIII&#44; Madrid&#44; Spain&#46; <span class="elsevierStyleSup"> 4</span>Unidad de Anatom&#237;a Patol&#243;gica&#44; Hospital Gregorio Mara&#241;&#243;n&#44; Madrid&#44; Spain&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Multiple endocrine neoplasia type 2A &#40;MEN2A&#41; is associated with specific germline missense mutations in the RET proto-oncogene&#46; It is an autosomal dominant trait with high penetrance and variable clinical expression&#46; Medullary thyroid carcinoma is the main clinical feature&#44; but&#46; there are variations&#44; even between members of the same family&#44; regarding the disease onset and its presentation&#46; Our objective was to explore whether two associated RET polymorphisms&#44; G691S and S904S&#44; could have any influence on the clinical form and the age at onset of the disease&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> G691S &#40;exon 11&#41; and S904S &#40;TCC-TCG&#44; exon 15&#41; polymorphisms of RET were analyzed in 198 individuals corresponding to 35 unrelated Spanish MEN2A families &#40;104 patients with oncogenic MEN 2A mutation and 94 healthy relatives&#41; and in a control population of 653 healthy individuals by amplification and sequencing analysis&#46; In all cases&#44; both polymorphisms co-segregated and were considered as a single variable in subsequent analyses&#46; The prevalence of G691S&#47;S904S polymorphisms was compared in MEN2A cases and their healthy relatives using the corrected Pearson&#39;s chi-square test allowing for correlation between members of the same family&#46; In the same way&#44; a possible correlation among cases between these polymorphisms and type of clinical presentation was assessed&#46; The relationship between G691S&#47;S904S polymorphisms and age at diagnosis in MEN2A patients was investigated considering &#34;age&#34; as a continuous variable and also as a dichotomous one&#44; taking 20 years as the cut-off&#46; Differences across G691S&#47;S904S groups were quantified using linear regression and logistic regression&#46; In both instances&#44; robust estimators of variance were used&#44; clustered in families&#46; The same analysis was restricted to index cases or probands&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> The studied polymorphisms followed Hardy-Weinberg equilibrium in the control population&#46; Among cases&#44; they were not related with the type of clinical presentation&#44; but homozygous were&#44; on average&#44; ten years younger when they were diagnosed &#40;p &#61; 0&#46;037&#41;&#46; In fact&#44; homozygous had an 8-fold probability to be diagnosed at an age before 20 &#40;p &#61; 0&#46;010&#41;&#46; Obviously&#44; these results could be biased given that the clinical diagnosis for some members of the same family may be conditioned on the time of diagnosis of the corresponding proband&#46; However&#44; when we focused specifically on index cases&#44; the association between age at onset and homozygote G691S &#47; S904S genotype persisted &#40;p &#60; 0&#46;001&#41;&#44; and the OR for being diagnosed before 20 was even stronger &#40;OR &#61; 19&#46;3&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusion&#58;</span> These results suggest that the presence of the RET polymorphisms G691S&#47;S904S seems to act as a genetic modifier causing an early appearance of the disease in MEN2A patients&#46; They could be used as markers in asymptomatic children of MEN2A families guiding time of surgical preventive resection&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">115</span><span class="elsevierStyleBold">GASTRIC CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREA IN SPAIN&#44; 1975-2000</span></p><p class="elsevierStylePara"> Nuria Aragon&#233;s&#44; Gonzalo L&#243;pez-Abente&#44; Marina Poll&#225;n&#44; Beatriz P&#233;rez-G&#243;mez&#44; Valent&#237;n Hern&#225;ndez&#44; Mario C&#225;rdaba&#44; Berta Su&#225;rez&#44; Alicia Estirado</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Centro Nacional de Epidemiolog&#237;a&#44; Instituto de Salud Carlos III&#44; Madrid&#44; Spain&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Though gastric cancer mortality has been declining during the last decades in Spain&#44; its evolution might have not been uniform across the country&#46; The objective of this analysis is to describe gastric cancer mortality trends by sex and geographical area in Spain&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material and methods&#58;</span> Both mortality and population data were obtained from the National Statistics Institute&#46; During the study period&#44; three revisions of the International Classification of Diseases were used &#40;8th&#44; 9th and 10th&#41;&#46; According to it&#44; 151-code was considered from 1975 to 1998 &#40;8th-9thICD&#41; and C16-code from 1999 to 2000 &#40;10thICD&#41;&#46; Individual records broken down by sex&#44; age group&#44; year of death and province of residence were used to compute age-adjusted mortality rates &#40;European standard population&#41; by sex&#44; year&#44; and Autonomous Community&#46; Joinpoint regression analysis was used to detect changes in trends from 1975 to 2000&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> In men&#44; Spain had an overall annual reduction of 3&#37; in gastric cancer mortality from 1975 to 2000&#46; Rates decreased a 4&#46;57&#37; per year until 1982&#44; where the speed of this decline was significantly reduced to 2&#46;56&#37;&#46; In women&#44; the same phenomenon was observed&#58; gastric cancer mortality rates decreased 3&#46;84&#37; per year from 1975-2000&#44; although rates descended a 5&#46;95&#37; each year until 1980&#44; where the fall in mortality slowed down to a 3&#46;51&#37;&#46; Among Autonomous Communities&#44; joinpoint regression analysis did not detect significant changes in trends for most of them neither in men nor in women&#46; There are&#44; however&#44; Autonomous Communities where gastric cancer rates for men stopped decreasing or even increased in recent years&#44; as Asturias&#44; Cantabria or Murcia&#46; When studying age groups from 35-64 and 65&#43;&#44; results were quite similar&#44; with annual changes in both sexes around -3&#46;0 and a statistically significant variation in trend around the early 80&#39;s in both age groups&#46; This similarity in trends among different age groups points toward a homogeneous cohort effect for successive generations&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> Gastric cancer mortality rates fell down in Spain from 1975 to 2000&#46; There is a significant steady decline from 1975 to the early 80&#39;s followed by a less accentuated reduction afterwards&#46; This general pattern of decrease could be observed in men and women of different age groups in most geographical areas&#46; There were nevertheless important differences in the magnitude of rates among geographical areas which persisted until recent years&#44; although these differences between Autonomous Communities tended to diminish along time&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">116</span><span class="elsevierStyleBold">PREVALENCIA DE VPH Y OTROS FACTORES DE RIESGO PARA LESIONES NEOPL&#193;SICAS PREINVASORAS</span></p><p class="elsevierStylePara"> Mireia Diaz Sanchis&#42;&#44; &#192;ngela Twose L&#225;zaro&#42;&#44; Jordi Ponce Sebasti&#224;&#42;&#42;&#44; M&#46; Dolores Mart&#237; Cardona&#42;&#42;&#44; Silvia de Sanjose Llongueres&#42;&#44; F&#46; Xavier Bosch i Jos&#233;&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Servicio de Epidemiolog&#237;a y Registro del C&#225;ncer&#44; Institut Catal&#224; d&#39;Oncologia&#44; Hospitalet de Llobregat&#44; Espa&#241;a&#46; &#42;&#42;Servicio de Ginecolog&#237;a&#44; Patolog&#237;a Cervical y Colposcop&#237;a&#44; Ciudad Sanitaria y Universitaria de Bellvitge&#44; Hospitalet de Llobregat&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introducci&#243;n&#58;</span> Las cl&#237;nicas de colposcop&#237;a asociadas a programas de cribaje reciben pacientes remitidas por un espectro de diagn&#243;sticos citol&#243;gicos con variabilidad de lectura importante&#46; La determinaci&#243;n del ADN de VPH puede contribuir a clarificar el pron&#243;stico de las lesiones ambiguas &#40;ASCUS&#41; y quiz&#225;s de las lesiones de bajo grado &#40;LSIL&#41;&#46; Los factores de riesgo convencionales de las lesiones preinvasoras tienen un bajo poder discriminatorio y escasa utilidad pron&#243;stica y de control cl&#237;nico&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">M&#233;todos&#58;</span> Se ha realizado un estudio caso-control entre las mujeres remitidas a una Unidad de Patolog&#237;a Cervical y Colposcop&#237;a por citolog&#237;a de cribado patol&#243;gica &#40;ASCUS&#44; LSIL&#44; HSIL&#41;&#46; A todas las mujeres se les realiz&#243; una entrevista epidemiol&#243;gica sobre factores de riesgo para neoplasia cervical&#44; una citolog&#237;a y biopsia de verificaci&#243;n y se obtuvo una muestra cervical para la determinaci&#243;n de VPH de alto riego mediante Captura de H&#237;bridos II&#46; En el an&#225;lisis estad&#237;stico se compararon mujeres con lesiones intraepiteliales de bajo y alto grado versus mujeres con lesiones de significado indeterminado mediante modelos de regresi&#243;n log&#237;stica polit&#243;mica&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> Hasta el momento se han reclutado 375 mujeres con citolog&#237;a de cribado patol&#243;gica distribuidas en ASCUS 18&#46;1&#37;&#44; lesi&#243;n intraepitelial de bajo grado &#40;LSIL&#41; 45&#44;1&#37; y lesi&#243;n intraepitelial de alto grado &#40;HSIL&#41; 36&#44;8&#37;&#46; El determinante mayor de las lesiones de bajo y alto grado versus lesiones ASCUS es la presencia de VPH HR&#44; la prevalencia es de 21&#44;5&#37; para las mujeres con ASCUS&#44; 44&#44;0&#37; para LSIL &#40;OR<span class="elsevierStyleInf">LSIL</span> &#61; 2&#44;4 &#40;1&#44;2-4&#44;8&#41;&#41; y 81&#44;8&#37; para HSIL &#40;OR<span class="elsevierStyleInf">HSIL</span> &#61; 15&#44;3 &#40;7&#44;2-32&#44;7&#41;&#41;&#46; Se detecta un aumento de riesgo asociado a haber tenido 4 o m&#225;s compa&#241;eros sexuales &#40;OR<span class="elsevierStyleInf">LSIL</span> &#61; 3&#44;3 &#40;1&#44;2-9&#44;2&#41;&#59; OR<span class="elsevierStyleInf">HSIL</span> &#61; 5&#44;0 &#40;1&#44;7-14&#44;4&#41;&#41; y al consumo de tabaco &#40;OR<span class="elsevierStyleInf">LSIL</span> &#61; 2&#44;6 &#40;1&#44;3-5&#44;6&#41;&#59; OR<span class="elsevierStyleInf">HSIL</span> &#61; 3&#44;6 &#40;1&#44;6-8&#44;1&#41;&#41;&#46; Aparece un efecto protector con 3 o m&#225;s partos &#40;OR<span class="elsevierStyleInf">LSIL</span> &#61; 0&#44;2 &#40;0&#44;1-0&#44;8&#41;&#59; OR<span class="elsevierStyleInf">HSIL</span> &#61; 0&#44;2 &#40;0&#44;1-0&#44;9&#41;&#41;&#46; Solamente para HSIL&#44; se halla riesgo asociado a haber tenido relaciones sexuales con compa&#241;eros casuales &#40;OR<span class="elsevierStyleInf">HSIL</span> &#61; 3&#44;0 &#40;1&#44;3-6&#44;9&#41;&#41; y un efecto protector con el uso de preservativos &#40;OR<span class="elsevierStyleInf">HSIL</span> &#61; 0&#44;5 &#40;0&#44;2-0&#44;9&#41;&#41;&#46; Ajustando por VPH desaparece el efecto del n&#250;mero de relaciones sexuales y el uso de preservativo&#44; pero se mantienen el consumo de tabaco&#44; el n&#250;mero de partos y para HSIL&#44; las relaciones sexuales con compa&#241;eros casuales&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> La prevalencia de ADN de VPH en lesiones citol&#243;gicas es sensiblemente distinta a la encontrada en estudios realizados en otras poblaciones y que se han establecido como referencia internacional&#46; Por ejemplo&#44; el estudio ALTS<span class="elsevierStyleSup">&#42;</span> describe ADN de VPH en 50&#37; de los ASCUS y en 80&#37; de las lesiones LSIL&#44; indicando la variabilidad y la especificidad local en la lectura de la citolog&#237;a&#46; Los resultados de referencia en estudios de triaje deben interpretarse en relaci&#243;n a las caracter&#237;sticas de los diagn&#243;sticos citol&#243;gicos locales&#46; La introducci&#243;n del test de VPH en cl&#237;nicas de colposcop&#237;a puede ayudar a establecer el pron&#243;stico y la conducta terap&#233;utica con mayor precisi&#243;n que la investigaci&#243;n sobre factores de riesgo de la paciente&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleSup">&#42;</span>J Natl Cancer Inst&#59;92&#58;397-402</p><p class="elsevierStylePara"><span class="elsevierStyleBold">117</span><span class="elsevierStyleBold">AN&#193;LISIS DE LA VARIACI&#211;N GEOGR&#193;FICA DE LA MORTALIDAD POR C&#193;NCER DE EST&#211;MAGO EN GALICIA</span></p><p class="elsevierStylePara"> Elisa Mar&#237;a Molanes&#44; M&#170; Eugenia Lado</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Servicio de informaci&#243;n sobre Sa&#250;de P&#250;blica&#44; Conseller&#237;a de Sanidade&#44; Santiago de Compostela&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introducci&#243;n&#58;</span> En los &#250;ltimos a&#241;os&#44; han surgido diferentes m&#233;todos estad&#237;sticos para estimar y suavizar las razones de mortalidad est&#225;ndar &#40;RME&#41;&#44; y poder as&#237; realizar comparaciones m&#225;s fiables del estado de salud de diferentes &#225;reas geogr&#225;ficas&#46; Los principales objetivos de este estudio son&#58; &#40;1&#41; aplicar un &#34;nuevo&#34; m&#233;todo bayesiano no param&#233;trico de suavizado de tasas y &#40;2&#41; comparar los resultados con los obtenidos al aplicar el modelo cl&#225;sico de Besag&#44; York y Molli&#233; &#40;modelo de convoluci&#243;n&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">M&#233;todos&#58;</span> Para llevar a cabo este estudio se obtuvieron del Registro de Mortalidad de Galicia&#44; los datos de mortalidad por c&#225;ncer de est&#243;mago &#40;CIE-9 151 y CIE-10 C16&#41; de la poblaci&#243;n masculina de Galicia en el per&#237;odo 1995-1999&#46; Se utilizaron las poblaciones a 1 de enero para cada a&#241;o&#44; municipio y grupo quinquenal de edad&#44; y como poblaci&#243;n est&#225;ndar se consider&#243; la poblaci&#243;n gallega en el per&#237;odo de estudio&#46; Con estos datos se calcularon las RME crudas y se estimaron las RME con ambos modelos&#46; Para ajustar el modelo bayesiano no param&#233;trico se utiliz&#243; el software BDCD y para el modelo de convoluci&#243;n el WinBUGS&#46; Para cada m&#233;todo se representaron geogr&#225;ficamente las estimaciones de las RME obtenidas para cada municipio&#44; as&#237; como su significaci&#243;n estad&#237;stica&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> La distribuci&#243;n espacial de las RME crudas no define claramente ninguna zona de riesgo de mortalidad por c&#225;ncer de est&#243;mago&#46; En el mapa obtenido con el modelo bayesiano no param&#233;trico se detect&#243; como zona con mayor riesgo de mortalidad toda la zona occidental de Galicia y la zona centro de la provincia de Lugo&#46; Sin embargo&#44; al estudiar su significaci&#243;n estad&#237;stica &#250;nicamente se mantuvo como &#225;rea de alto riesgo parte de la zona occidental de Galicia&#46; En lo que se refiere al mapa basado en el modelo de convoluci&#243;n&#44; se detect&#243; como &#250;nica zona de riesgo la parte occidental de Galicia y&#44; asimismo&#44; s&#243;lo una parte de ella se mantuvo como &#225;rea de riesgo en su mapa de significaci&#243;n&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> De este estudio se concluye que los mapas que representan las RME estimadas son menos ruidosos que el que representa las RME crudas y que ambos modelos coinciden en detectar como zona de mayor riesgo de mortalidad por c&#225;ncer de est&#243;mago la costa occidental de Galicia&#46; Aunque los resultados obtenidos son muy similares&#44; el modelo bayesiano no param&#233;trico se muestra m&#225;s adecuado que el modelo de convoluci&#243;n para la representaci&#243;n geogr&#225;fica del mapa de riesgos de &#225;reas peque&#241;as&#44; ya que adem&#225;s de suavizar el valor de las RME en cada &#225;rea permite detectar discontinuidades en el mapa&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">118</span><span class="elsevierStyleBold">INCIDENCIA Y SUPERVIVENCIA RELATIVA DE LOS LINFOMAS NO HODGKIN EN GIRONA 1994-1999</span></p><p class="elsevierStylePara"> Rafael Marcos Gragera&#42;&#44; &#192;ngel Izquierdo Font&#42;&#44; Cristalina Fern&#225;ndez Fidalgo&#42;&#42;&#44; Santiago Gardella&#42;&#42;&#44; M&#170; Loreto Vilardell Gil&#42;&#44; Maria Bux&#243; Pujolr&#224;s&#42;&#44; Pau Viladiu Quemada&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Unitat d&#39;Epidemiologia i Registre de C&#224;ncer de Girona&#44; Institut Catal&#224; d&#39;Oncologia de Girona&#44; Girona&#44; Spain&#46; &#42;&#42;Servei d&#39;Hematologia&#44; Instiut Catal&#224; d&#39;Oncologia de Girona&#44; Girona&#44; Spain&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Objetivo&#58;</span> Conocer la incidencia y supervivencia relativa poblacional de los linfomas no Hodgkin &#40;LNH&#41; en Girona&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material y m&#233;todos&#58;</span> A partir de los datos obtenidos por el Registro poblacional de C&#225;ncer de Girona&#44; se analiz&#243; la incidencia y supervivencia de los linfomas no Hodgkin en la Regi&#243;n Sanitaria Girona &#40;RSG&#41;&#46; La poblaci&#243;n cubierta por el registro seg&#250;n el censo de 1996 fue de 518&#46;531 habitantes&#46; Se calculan las tasas de incidencia brutas &#40;T&#46;B&#41; y ajustadas &#40;T&#46;Aj&#41; a la poblaci&#243;n est&#225;ndar mundial&#46; Para el c&#225;lculo de la supervivencia se hizo un seguimiento de los pacientes hasta 12&#46; 1999&#46; Se calcul&#243; la supervivencia relativa&#44; tasa entre la supervivencia observada y la esperada&#44; calculada &#233;sta en funci&#243;n de la mortalidad de la poblaci&#243;n de Girona&#46; Se utiliz&#243; el m&#233;todo de Est&#232;ve&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span></p><p class="elsevierStylePara"><img src="138v17nSupl.2-13051580tab01.gif"></img></p><p class="elsevierStylePara"><img src="138v17nSupl.2-13051580tab02.gif"></img></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> En la RSG los linfomas no Hodgkin ocupan el s&#233;ptimo lugar en orden de frecuencia&#44; tanto hombres como en mujeres&#46; La incidencia de los LNH es mas alta en los hombres&#46; Cuando comparamos con las cifras de incidencia obtenidas en el resto de registros espa&#241;oles &#40;EUROCIM&#44; ENCR&#41; observamos&#44; en el caso de los hombres&#44; una incidencia estad&#237;sticamente superior de los LNH en la RSG&#46; A nivel internacional la incidencia de los LNH en la RSG es situar&#237;a a un nivel intermedio - alto&#46; &#40;Cancer Incidence in Five Continents VII&#44; 1997&#41;&#46; La SR de los LNH es similar a la que se da en el resto de registros espa&#241;oles&#44; europeos y americanos &#40;EUCAN&#44; 1997&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">119</span><span class="elsevierStyleBold">LUNG CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREAS IN SPAIN&#46; 1975-2000</span></p><p class="elsevierStylePara"> Mario C&#225;rdaba&#44; Nuria Aragon&#233;s&#44; Marina Poll&#225;n&#44; Beatr&#237;z P&#233;rez&#44; Berta Su&#225;rez&#44; Alicia Estirado&#44; Valent&#237;n Hern&#225;ndez&#44; Gonzalo L&#243;pez-Abente</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Epidemiolog&#237;a Ambiental y C&#225;ncer&#44; Instituto de Salud Carlos III&#44; Madrid&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Lung cancer mortality has been declining since 1990s in Europe&#46; Spain presents a delayed pattern due to its different phase of tobacco epidemic&#46; The aim of this analysis is to describe lung cancer mortality trends by sex&#44; age and geographical area in Spain&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material and methods&#58;</span> Lung cancer mortality &#40;International Classification of Diseases &#40;ICD&#41;8th&#58;-9th&#58;162&#59; ICD-10th&#58;C34&#41; and population data were obtained from the National Statistics Institute&#46; Individual records broken down by sex&#44; age&#44; year of death and province of residence were used to compute age-adjusted and age-adjusted truncated &#40;35-64&#41; rates &#40;European standard population&#41; by sex&#44; year&#44; and Autonomous Community&#46; Joinpoint regression analysis was used to detect changes in trends between 1975-2000&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> Age-adjusted rates for Spanish males showed an annual percent of change&#40;APC&#41; for 1975-2000 of 2&#46;11&#37;&#46; However&#44; joinpoint analysis detected changes in trends in 1988 and 1994 &#40;1975-1988 APC&#58;3&#46;54&#37;&#59; 1988-1994&#58;1&#46;58&#37;&#59; 1994-2000&#58; -0&#46;35&#37;&#41;&#46; Truncated rates showed an increase of 2&#46;05&#37; between 1975-200&#44; although rates increased a 3&#46;3&#37; per year until 1990 and reached a plateau since&#46; Among regions&#44; global rates show a positive APC for period 1975-2000 ranging from 0&#46;6&#37; in Cantabria up to 3&#46;52&#37; in Castile-La Mancha&#46; Truncated rates follow a similar homogeneous pattern&#46; Joinpoint analysis detected significant changes in trends in late 1980s or early 1990s&#44; changing from a clear increase to smooth increments&#44; plateaus or even a decline in rates afterwards&#46; Spanish females&#44; for global rates&#44; showed an annual increase of 0&#46;47&#37; between 1975-2000&#46;Nevertheless&#44; joinpoint analysis detected a change in trend in 1990&#44; moving from an annual decrease of -0&#46;71&#37; to an increase of 2&#46;39&#37;&#46; Truncated rates presented an increase of 1&#46;15&#37; for period 1975-2000&#44; but again an acute change in trend turns up in 1990&#44; when APC shifted sharply from -1&#46;43&#37; to 5&#46;38&#37; annually&#46; Among regions&#44; global rates between 1975-2000 seem irregular&#46; Some suggest a rising slope &#40;Balearic Islands&#58;1&#46;53&#37;&#59; Madrid&#58;2&#46;52&#37;&#59; Basque Country&#58;1&#46;87&#37;&#41;&#46; Others show a minimal increase&#40;Catalonia&#58; 0&#46;09&#37;&#59; Castile-Leon&#58; 0&#46;24&#37;&#41; or even a decline &#40;Extremadura&#58; -1&#46;75&#37;&#44; La Rioja&#58; -1&#46;24&#37;&#44; Aragon&#58;-0&#46;65&#37;&#41;&#46; Truncated rates exhibit a similar pattern though in general increases are greater&#46; Joinpoint analysis detected changes in global and truncated rates&#46; Some regions imitate the nationwide pattern of decrease-increase while others do not&#46; These changes came about in late 1980s or middle 1990s and are sharper in truncated rates&#46; However&#44; Extremadura and Castile-La Mancha present a continuous decline between 1975-2000&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> While time trends in lung cancer mortality in men are levelling off since 1990s&#44; in females they are increasing sharply&#44; specially in 35-64 age group&#44; pointing to the beginning of the epidemic phenomenon of lung cancer in women that is affecting to cohorts born after 1940&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">120</span><span class="elsevierStyleBold">FEMALE BREAST CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREAS IN SPAIN&#46; 1975-2000</span></p><p class="elsevierStylePara"> Berta Su&#225;rez&#44; Nuria Aragon&#233;s&#44; Marina Poll&#225;n&#44; Beatriz P&#233;rez-G&#243;mez&#44; Mario C&#225;rdaba&#44; Alicia Estirado&#44; Valent&#237;n Hern&#225;ndez&#44; Gonzalo L&#243;pez-Abente</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Centro Nacional de Epidemiolog&#237;a&#44; Instituto de Salud Carlos III&#44; Madrid&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Breast cancer is the commonest malignancy in women in Spain&#46; Though mortality rates have decreased during the last decade&#44; within the country each region has had a different pattern of mortality&#46; The objective of this investigation is to describe female breast cancer mortality trends by geographical area in Spain&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material and methods&#58;</span> Female breast cancer mortality &#40;International Classification of Diseases &#40;ICD 8th-9th&#58;174&#59; ICD-10th&#58;C50&#41; and population data were obtained from the National Statistics Institute&#46; Individual records broken down by age group&#44; year of death and province of residence were used to compute age-adjusted mortality rates &#40;European Standard population&#41; and age-adjusted truncated rates &#40;35-64 years&#41; by year and by periods of five years for each Autonomous Community&#40;AC&#41; and for Spain as a whole&#46; Joinpoint regression analysis was used to detect statistically significant changes in trends from 1975 to 2000&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> In Spain&#44; age-adjusted mortality rates increased until the 1991-1995 period&#46; Joinpoint analysis detected the change in trend in 1991&#58; while the annual increase until this year was 2&#46;41&#37;&#44; rates begun to decline since &#40;-1&#46;75&#37;&#41;&#46; Truncated rates had a quite similar evolution&#44; though in this case the annual decrease from 1992 onwards was bigger &#40;-3&#46;33&#37;&#41;&#46; This initial increase in rates&#44; followed by a descending trend&#44; was found across the whole country&#44; excepting Cantabria where mortality rates had a continuous light increase&#46; Joinpoint analysis detected statistically significant changes in trend between 1990-95 in most of the Autonomous Communities&#46; Although apparently rates in Aragon&#44; Valencian Community and Basque Country showed a similar trend&#44; changes were not statistically significant&#46; La Rioja had the highest increase&#44; 3&#46;42&#37; per year&#44; and a very fast decline&#44; -6&#46;35&#37;&#46; It is noteworthy the intense decrease observed in Navarra &#40;-8&#44;28&#37;&#41; while in the other ACs descents ranged between -1&#44;04 in Murcia and -3&#44;4&#37; in Catalonia&#46; Also remarkable was Madrid&#44; where two points of change were identified&#44; defining three different periods&#58; rates had a slow increase until 1985 &#40;1&#46;08&#37;&#41;&#44; a sharp ascent till 1988 &#40;12&#46;74&#37;&#41; and a slight decline since that year &#40;1&#46;04&#37;&#41;&#46; Truncated rates trends among Autonomous Communities had a similar evolution than all-ages groups rates&#44; though with a higher decreasing slope&#46; Every community had a change in trend but the joinpoint analysis only detected significant changes in 10 of them&#46; Murcia was the community with the highest increase before the joinpoint &#40;3&#46;05&#37;&#41;&#44; followed up by La Rioja &#40;2&#46;96&#37;&#41; whose decreasing slope afterwards was very high &#40;-9&#46;33&#37;&#41;&#46; Madrid had again a different pattern from the other ACs&#44; similar to the one observed for all-ages groups rates&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> Breast cancer mortality rates increased in Spain between 1975-1991 and then declined&#46; Screening programs and the improvement in early diagnostic and therapeutic methods might explain this important decline in mortality rates among women&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">121</span><span class="elsevierStyleBold">NON-HODGKIN LYMPHOMAS MORTALITY IN MADRID</span></p><p class="elsevierStylePara"> Berta Su&#225;rez&#42;&#44; Gonzalo L&#243;pez-Abente&#42;&#44; Consuelo Ib&#225;&#241;ez&#42;&#42;&#44; Valent&#237;n Hern&#225;ndez&#42;&#44; Mario C&#225;rdaba&#42;&#44; Alicia Estirado&#42;&#44; Nuria Aragon&#233;s&#42;&#44; Beatriz P&#233;rez-G&#243;mez&#42;&#44; Marina Poll&#225;n&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;National Centre for Epidemiology&#44; Carlos III Institute&#44; Madrid&#44; Spain&#46; &#42;&#42;Epidemiology Unit&#44; Madrid Regional Health Authority&#44; Madrid&#44; Spain&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> To study the spatial pattern of home addresses of deaths by non-hodgkin lymphomas in the Madrid Autonomous Community&#44; considering mortality as a spatial point process&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material and methods&#58;</span> The mortality registry of the Statistics Institute of the Community of Madrid supplied a database containing information from all death certificates with mention to non-hodgkin lymphomas &#40;NHL&#41; during the period 1991-1997as well as a randomly selected sample of 1500 controls&#44; stratified by year of death and sex&#46; All permanent home addresses of cases and controls were georeferencied in UTM-coordinates&#46; Spatial clusters were detected by means of an approach based on the study of Ripley&#39;s K functions differences among cases and controls&#46; In order to identify clusters&#44; we obtained a relative risk surface comparing the kernel-smoothed spatial intensity of the process among cases and controls&#46; Its intersection with tolerance bounds from the constant region wide relative risk hypothesis allow to locate the clusters&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> 1502 cases of NHL were registered&#46; The case-control comparison showed a possible borderline significant cluster around a distance of 200 meters&#46; The study of the spatial intensity allows to identify NHL clusters in the south area of Torrej&#243;n de Ardoz and in two districts of Madrid city&#46; A classical analysis by logistic regression found the following results for the municipal covariate&#58; Torrej&#243;n OR&#61;1&#46;96 &#40;95&#37; CI 0&#46;76-1&#46;08&#41; and Madrid OR&#61;1&#46;20 &#40;95&#37; CI 0&#46;99-1&#46;45&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> The study of mortality as a spatial point process may be an useful tool to detect patterns that could remain hidden with lattice data analysis&#46; The identified spatial NHL clusters could match partially with the distribution of AIDS rates in the Madrid Region&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">122</span><span class="elsevierStyleBold">EVOLUCI&#211;N DE LA MORTALIDAD POR CANCER DE MAMA EN CATALUNYA&#44; 1991-2000</span></p><p class="elsevierStylePara"> Xavi Puig&#44; Rosa Gispert&#44; Anna Puigdef&#224;bregas</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Servei d&#39;Informaci&#243; i Estudis&#44; Departament de Sanitat i Seguretat Social&#44; Barcelona&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Antecedentes&#58;</span> La evoluci&#243;n de la mortalidad por c&#225;ncer de mama en a&#241;os recientes manifiesta una clara inflexi&#243;n en la tendencia al aumento significativos mostrado en a&#241;os anteriores&#46; En esta evoluci&#243;n pueden influir tanto la incidencia de este tumor como su supervivencia&#44; aspecto relacionado con la efectividad de la asistencia sanitaria que se presta a estas pacientes&#46; El objetivo del trabajo es analizar si la evoluci&#243;n de la mortalidad por c&#225;ncer de mama por grupo de edad es consistente con la tendencia general de esa causa de mortalidad en Catalunya&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">M&#233;todos&#58;</span> Se han empleado las defunciones por c&#225;ncer de mama &#40;CIM-9&#58;174&#59; CIM-10&#58;C50&#41; del per&#237;odo 1991-2000 del Registro de Mortalidad del Departament de Sanitat de Catalunya&#44; y la poblaci&#243;n a partir de estimaciones intercensales y postcensales&#46; Los datos se disponen truncados para las mujeres mayores de 34 a&#241;os&#44; y estratificados por grupos de edad decenales&#44; siendo el &#250;ltimo abierto 85 a&#241;os y m&#225;s&#46; Para evaluar la tendencia se han ajustado modelos de poisson para cada grupo de edad&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> En el per&#237;odo 1991-2000 se registraron 10&#46;116 defunciones por c&#225;ncer de mama en mujeres mayores de 34 a&#241;os&#46; El porcentaje global de cambio anual es de -2&#44;7&#37;&#46; Esta evoluci&#243;n no es homog&#233;nea por grupos de edad&#44; en los que se observa un claro gradiente&#44; as&#237; las mujeres de 35 a 44 a&#241;os son las que han experimentado una reducci&#243;n m&#225;s acusada&#44; del -5&#44;6&#37;&#44; y paulatinamente se modera la tendencia descendiente con la edad&#44; siendo en las mujeres mayores de 84 a&#241;os de -0&#44;5&#37;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> La reducci&#243;n de las tasas de mortalidad por c&#225;ncer de mama a diferentes edades ha sido muy importante en la &#250;ltima d&#233;cada&#44; con un manifiesto gradiente relacionado con la edad&#46; La reducci&#243;n del la mortalidad de manera consistente en todos los grupos de edad suscribe el efecto beneficioso de las intervenciones sanitarias frente a este tumor&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">123</span><span class="elsevierStyleBold">CHEMICALS AND ELECTROMAGNETIC FIELDS OCCUPATIONAL EXPOSURE AND RISK OF TESTICULAR CANCER AMONG SWEDISH MEN</span></p><p class="elsevierStylePara"> Alicia Estirado&#42;&#44; Marina Poll&#225;n&#42;&#44; Beatriz P&#233;rez-G&#243;mez&#42;&#44; Per Gustavsson&#42;&#42;&#44; Nils Plato&#42;&#42;&#44; Girgitta Floderus&#42;&#42;&#42;&#44; Nuria Aragon&#233;s&#42;&#44; Montse Alcalde&#42;&#44; Gonzalo L&#243;pez-Abente&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;National Centre for Epidemiology&#44; Carlos III Institute of Health&#44; Madrid&#44; Spain&#46; &#42;&#42;Department of Public Health Sciences&#44; Karolinska Institutet&#44; Stockholm&#44; Sweden&#46; &#42;&#42;&#42;Institute of Environmental Epidemiology&#44; Karolinska Institutet&#44; Stockholm&#44; Sweden&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Previous studies have reported an association between testicular cancer and some jobs&#44; suggesting that certain occupational exposures could play an etiological role&#46; The aim of this study is the association between seminoma and nonseminoma tumours and the occupational exposure to chemicals and electromagnetic fields &#40;ELMF&#41; among Swedish men&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Material and methods&#58;</span> The base population for this historical cohort comprised all Swedish men recorded in the 1960 census&#44; who were gainfully employed at the time of the 1970 census&#44; and were still alive and over the age of 24 years on January 1&#44; 1971&#46; The follow-up period was 19 years &#40;1971-1989&#41;&#46; The Swedish cancer environmental register was used to compute specific rate numerators&#44; and the 1970 census to compute specific rate denominators&#46; Exposure to 13 chemical factors was assessed by linking each combination of occupation and industrial branch to a Swedish job-exposure matrix &#40;JEM&#41;&#44; which classifies them as probable&#44; possible and non exposed&#46; Exposure to ELMF was assessed using a Swedish JEM based on the 100 most common jobs among men&#46; The interaction between chemicals and ELMF was done in the subcohort of subjects with information available for both exposures&#46; Relative risks &#40;RRs&#41; adjusted for age&#44; period&#44; geographical area&#44; town-size and occupational sector were computed using log-linear Poisson models&#46; The same analyses were repeated for young people &#40;&#60;40 years&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> During follow-up a total of 748 seminomas and 405 nonseminomas were reported&#46; In the general cohort&#44; 39 seminomas were possibly exposed to peak of pesticides&#44; with RR&#58; 1&#46;17 &#40;CI 95&#37;&#58; 0&#46;70 - 1&#46;94&#41;&#46; While exposure to solvents for nonseminomas presented a dose response relationship with RR 1&#46;17 for possible exposed and 1&#46;21 for probable exposed&#44; none of them attained statistical significance&#46; Only 8 nonseminomas were possibly exposed to petroleum products&#58; RR&#58; 1&#46;34 &#40;CI 95&#37;&#58; 0&#46;65 - 2&#46;77&#41;&#44; and none was probably exposed to this product&#46; In summary&#44; no statistically significant association was found between chemicals or ELMF and testicular cancer in the general cohort or in the subgroup of younger workers&#46; There was not observed an interaction between ELMF and any of the chemicals studied&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> Our results did not corroborate the previously reported increased risks for occupational exposure to solvents&#44; oil mixtures&#44; petroleum products&#44; PAH or metals&#46; Nevertheless&#44; exposure misclassification caused by JEM may have biased the RRs towards the null hypothesis&#46; ELMF did not act as a risk factor or as an effect modifier for testicular cancer in this cohort&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">124</span><span class="elsevierStyleBold">C&#193;NCER DE MAMA EN LA PROVINCIA DE C&#193;DIZ&#58; VARIABLES SOCIODEMOGR&#193;FICAS&#44; CONDUCTAS EN SALUD Y ESTADIO AL DIAGN&#211;STICO</span></p><p class="elsevierStylePara"> Mar&#237;a Victoria Garc&#237;a-Palacios&#42;&#44; Encarnaci&#243;n Ben&#237;tez&#42;&#44; Soledad M&#225;rquez&#42;&#42;&#44; Antonio Escolar&#42;&#44; Estrella Figueroa&#42;&#42;&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Servicio de Medicina Preventiva y Salud P&#250;blica&#44; Hospital Universitario Puerta del Mar&#44; C&#225;diz&#44; Espa&#241;a&#46; &#42;&#42;Escuela Andaluza de Salud P&#250;blica&#44; Granada&#44; Espa&#241;a&#46; &#42;&#42;&#42;Servicio de Medicina Preventiva y Salud P&#250;blica&#44; Hospital Universitario de Puerto Real&#44; C&#225;diz&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Objetivos&#58;</span> Describir los factores sociodemogr&#225;ficos&#44; conductas en salud y estadio posquir&#250;rgico TNM de las mujeres diagnosticadas de c&#225;ncer de mama&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Metodolog&#237;a&#58;</span> El estudio prospectivo incluy&#243; todos los casos incidentes de tumor maligno de mama con confirmaci&#243;n histopatol&#243;gica entre el 1 Enero 1999 y 30 Junio 2000 en mujeres residentes en las &#225;reas hospitalarias de C&#225;diz y Puerto Real&#46; Mediante entrevista por personal entrenado se midieron variables sociodemogr&#225;ficas &#40;edad&#44; estado civil&#44; nivel de educaci&#243;n&#44; municipio de residencia&#41;&#44; cl&#237;nicas &#40;primer s&#237;ntoma manifestado y n&#250;mero total de s&#237;ntomas&#41;&#44; del sistema sanitario &#40;distrito&#44; cobertura&#44; tipo de consulta ante el primer s&#237;ntoma&#41;&#44;conductas en salud de la mujer &#40;participaci&#243;n en actividades de detecci&#243;n precoz&#44; autoexploraci&#243;n&#41; y tiempo transcurrido hasta el diagn&#243;stico &#40;retraso debido al paciente&#58; tiempo transcurrido desde la aparici&#243;n del primer s&#237;ntoma a la visita a un m&#233;dico y retraso debido al sistema&#58; desde la visita a un m&#233;dico al diagn&#243;stico&#41;&#46; El estadio tumoral en el momento del diagn&#243;stico se midi&#243; mediante clasificaci&#243;n TNM&#46; Se realiz&#243; an&#225;lisis de frecuencias&#44; medias&#44; DE y porcentajes para la descripci&#243;n de los datos&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> Se estudiaron 179 pacientes &#40;40 asintom&#225;ticas diagnosticadas por screening&#41; con una edad media de 56&#44;9 a&#241;os &#40;rango&#58; 26-90&#41; siendo el 44&#44;4&#37; mujeres &#60; 50 a&#241;os&#46; La mayor parte &#40;59&#44;2&#37;&#41; no ten&#237;a estudios primarios completos&#46; El 60&#37; de las mujeres present&#243; un solo s&#237;ntoma&#44; siendo el m&#225;s frecuente el bulto mamario&#46; Un 46&#37; de las mujeres con edad en criterio de screening no acudi&#243; al mismo&#46; En relaci&#243;n a la autoexploraci&#243;n&#44; el 74&#37; no hab&#237;a o&#237;do hablar de este t&#233;rmino&#44; aunque la mitad de las mujeres la practicaban habitualmente &#40;55&#44;5&#37;&#41;&#46; El 60&#37; de las pacientes acudi&#243; al m&#233;dico en el primer mes desde la percepci&#243;n de su primer s&#237;ntoma&#46; &#40;media 75&#44;4&#59; DE 184&#44;35&#59; mediana 8&#44;0&#59; rango 0-1106 d&#237;as&#41;&#46; El tratamiento quir&#250;rgico se realiz&#243; tras una media de 96&#44;32 d&#237;as &#40;DE 130&#44;18&#44; mediana 55&#44;0&#59; rango 2-1045&#41; siendo el estadio posquir&#250;rgico m&#225;s frecuentemente encontrado el IIa y IIb&#46; El 68&#37; present&#243; un tumor &#62; 2 cent&#237;metros&#44; mientras que solamente el 4&#44;1&#37; present&#243; un estadio con tumor in situ&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> Las mujeres de este estudio tienen una edad similar a la descrita en la literatura&#44; y la mayor&#237;a tiene un nivel educativo bajo&#46; La participaci&#243;n en programas de screening fue menor a la considerada como necesaria para reducir la mortalidad por c&#225;ncer &#40;70&#37;&#41;&#44; por lo que se deben investigar formas de penetraci&#243;n en estos colectivos y otras posibles estrategias de prevenci&#243;n en las mujeres m&#225;s j&#243;venes&#44; no incluidas en estos programas y que representaron casi la mitad de los diagn&#243;sticos&#46; El tratamiento de estas pacientes sufri&#243; una demora sanitaria considerable siendo la mayor&#237;a diagnosticadas con un tumor &#62; 2 cent&#237;metros&#44; momento en el que la enfermedad comienza a crecer m&#225;s r&#225;pidamente&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">125</span><span class="elsevierStyleBold">FACTORES SOCIODEMOGR&#193;FICOS&#44; SANITARIOS&#44; CONDUCTA EN SALUD Y ESTADIO DIAGN&#211;STICO DEL C&#193;NCER DE MAMA&#58; COMPARACI&#211;N SEG&#218;N DOS CLASIFICACIONES</span></p><p class="elsevierStylePara"> Mar&#237;a Victoria Garc&#237;a-Palacios&#42;&#44; Encarnaci&#243;n Ben&#237;tez&#42;&#44; Soledad M&#225;rquez&#42;&#42;&#44; Antonio Escolar&#42;&#44; Estrella Figueroa&#42;&#42;&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Servicio de Medicina Preventiva y Salud P&#250;blica&#44; Hospital Universitario Puerta del Mar&#44; C&#225;diz&#44; Espa&#241;a&#46; &#42;&#42;Escuela Andaluza de Salud P&#250;blica&#44; Granada&#44; Espa&#241;a&#46; &#42;&#42;&#42;Servicio de Medicina Preventiva y Salud P&#250;blica&#44; Hospital Universitario de Puerto Real&#44; C&#225;diz&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Objetivos&#58;</span> Identificar si existe asociaci&#243;n entre los factores socioecon&#243;micos&#44; sanitarios y conductas en salud previas al diagn&#243;stico&#44; y el estadio posquir&#250;rgico TNM del c&#225;ncer de mama seg&#250;n dos estadiajes diferentes&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Metodolog&#237;a&#58;</span> El estudio prospectivo incluy&#243; todos los casos incidentes de c&#225;ncer de mama con confirmaci&#243;n histopatol&#243;gica entre el 1 Enero 1999 y 30 Junio 2000 en mujeres residentes en las &#225;reas hospitalarias de C&#225;diz y Puerto Real&#46; Se realiz&#243; entrevista personal midiendo factores de utilizaci&#243;n del sistema sanitario&#44; variables cl&#237;nicas&#44; participaci&#243;n en actividades de detecci&#243;n precoz&#44; autoexploraci&#243;n&#44; retraso debido al paciente y al sistema y otras variables &#40;edad&#44; estado civil&#44; nivel de estudios&#46;&#46;&#46;&#41;&#46; La variable dependiente TNM se reconvirti&#243; en estadios cl&#225;sicos &#40;I-IV&#41; y en un segundo estadiaje seg&#250;n diagn&#243;stico precoz &#40;precoz&#58; T0-is-T1&#44; N0-N1&#44; M0 y avanzado&#58; T2-T4&#44; N0-N1&#44;M1&#41;&#46; Mediante regresi&#243;n log&#237;stica se analiz&#243; la relaci&#243;n entre el estadio encontrado y las posibles variables predictoras&#46; Se ha evaluado el ajuste de los modelos&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> Se estudiaron 179 pacientes&#44; siendo los estadios cl&#225;sicos m&#225;s frecuentes el IIa &#40;33&#44;1&#37;&#41; y IIb &#40;21&#44;9&#37;&#41;&#59; presentando el 68&#37; un diagn&#243;stico no precoz&#46; Se encontr&#243; un mayor riesgo de presentar un tama&#241;o tumoral &#62; 2 cm en las mujeres no diagnosticadas por screening &#40;OR &#61; 2&#44;29&#41;&#46; De los 2 modelos de regresi&#243;n finales &#40;excluy&#233;ndose 40 mujeres asintom&#225;ticas&#41;&#44; se relacion&#243; un mayor riesgo de lesiones &#62; 2 cm &#40;diagn&#243;stico avanzado&#41; con&#58; mujeres no casadas &#40;OR &#61; 4&#44;00&#41;&#44; no realizar autoexploraci&#243;n &#40;OR&#61;2&#44;55&#41;&#44; presentar m&#225;s de un s&#237;ntoma &#40;OR &#61; 3&#44;89&#41; y mujeres cuya decisi&#243;n de consultar al m&#233;dico no fuese exclusivamente suya &#40;OR &#61; 2&#44;57&#41;&#46; Se asoci&#243; un mayor retraso en el sistema en los tumores m&#225;s peque&#241;os &#40;5 meses&#44; frente 3 meses en los de mayor tama&#241;o&#41;&#46; El modelo cl&#225;sico relacion&#243; con un mayor estadio&#58; un menor nivel de estudios &#40;OR&#61;7&#44;8&#41;&#44; no realizar autoexploraci&#243;n mamaria &#40;OR &#61; 2&#44;46&#41;&#44; tener un s&#237;ntoma diferente al bulto mamario &#40;OR &#61; 3&#44;73&#41; y cuando la decisi&#243;n de acudir al m&#233;dico no la tomaba exclusivamente la paciente &#40;OR &#61; 2&#44;66&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> La clasificaci&#243;n basada en un diagn&#243;stico precoz &#40;T &#60; 2 cm&#41;&#44; manifest&#243; la importancia del screening en mujeres asintom&#225;ticas y de la consulta precoz ante la presencia de un primer s&#237;ntoma&#44; para lograr diagn&#243;sticos tumorales menores y susceptibles de cirug&#237;a conservadora&#46; No obstante&#44; las pacientes con menor tama&#241;o tumoral sufrieron un mayor retraso en el sistema&#44; cuestionando la preparaci&#243;n de dispositivos para un tratamiento definitivo que amenazar&#237;a la utilidad de la detecci&#243;n precoz&#46; La conducta en salud de la mujer&#44; como la discutida autoexploraci&#243;n&#44; se relacion&#243; con un tama&#241;o tumoral menor y estadios menos avanzados&#46; El bajo nivel cultural y el presentar un s&#237;ntoma diferente al bulto mamario aparecen como el principal predictor de una enfermedad avanzada&#44; por lo que deber&#237;a plantearse una informaci&#243;n dirigida a estos grupos que evitara un retraso excesivo por parte del paciente&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">126</span><span class="elsevierStyleBold">BREAST CANCER RISK FACTORS&#44; ACCORDING TO JOINT ESTROGEN RECEPTOR&#47;PROGESTERONE RECEPTOR STATUS OF TUMOR</span></p><p class="elsevierStylePara"> Jennifer A&#46; Rusiecki&#44; Theodore R&#46; Holford&#44; Tongzhang Zheng</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Department of Epidemiology and Public Health&#44; Yale University&#44; School of Medicine&#44; New Haven&#44; USA&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Most of the risk factors identified for breast cancer have been found to have modest relative risks and account for only a portion of breast cancer cases&#46; Characterization of breast tumors on both estrogen receptor &#40;ER&#41; and progesterone receptor &#40;PR&#41; status suggests distinct biological and clinical profiles&#46; Therefore&#44; analyzing breast cancer as one disease may obscure associations with suspected risk factors and lead to studies finding weak associations&#46; Based on these profiles&#44; it has been hypothesized that tumors responsive for both hormones &#40;ER&#43;PR&#43;&#41; are most closely associated with hormone-related risk factors&#44; tumors unresponsive for both hormones &#40;ER-PR-&#41; are less associated with these risk factors&#44; and receptor discordant tumors &#40;ER&#43;PR- and ER-PR&#43;&#41; have intermediate effects&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> In this case-control study we investigated whether the effects of suspected risk factors for breast cancer varied by joint ER&#47;PR status of tumor&#44; by examining age at menarche&#44; age at first full term pregnancy&#44; nulliparity&#44; lifetime lactation&#44; menopausal status&#44; body mass index&#44; ever use of estrogen&#44; alcohol intake&#44; smoking&#44; family history and race&#44; for four tumor subtypes &#40;ER&#43;PR&#43;&#44; ER-PR-&#44; ER&#43;PR- and ER-PR&#43;&#41;&#46; For a given risk factor&#44; odds ratios with respect to the common control group were compared using multiple logistic regression&#44; adjusted for all other risk factors simultaneously&#46; We also compared the ER&#43;PR&#43; case group to the ER-PR- case group&#44; since it has been hypothesized that they represent the two breast cancer subtypes which differ most substantially&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> Among 420 cases and 406 controls&#44; the effects of some risk factors varied by joint ER&#47;PR status&#46; Early age at menarche &#40;&#60;12 years&#41; was most strongly associated with ER-PR&#43; tumors &#40;OR&#61;2&#46;3&#59; 95&#37;CI&#44; 0&#46;7-8&#46;4&#41;&#46; Additionally&#44; women with ER&#43;PR&#43; tumors were 2&#46;2 times more likely to have experienced early menarche than women with ER-PR- tumors &#40;95&#37;CI&#44; 0&#46;8-6&#46;2&#41;&#46; Women who had an older age at their first pregnancy &#40;&#61; 30 years&#41; were most likely to have ER&#43;PR- tumors &#40;OR&#61;2&#46;2&#59; 95&#37;CI&#44; 1&#46;1-4&#46;5&#41;&#46; Women who reported ever having consumed alcohol were 3&#46;4 times more likely to have ER&#43;PR&#43; tumors than ER-PR- tumors &#40;95&#37;CI&#44; 1&#46;4-8&#46;4&#41;&#46; Family history of breast cancer was most closely associated with ER&#43;PR&#43; tumors &#40;OR&#61;1&#46;5&#59; 95&#37;CI&#44; 0&#46;9-2&#46;5&#41; and ER&#43;PR- tumors &#40;OR&#61;1&#46;5&#59; 95&#37;CI&#44; 0&#46;8-2&#46;9&#41;&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> Although we detected several interesting individual effects&#44; there was no clear pattern of association whereby ER&#43;PR&#43; tumors were most closely associated with hormonally mediated risk factors and ER-PR- tumors were more closely associated with non-hormonally mediated risk factors&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">127</span><span class="elsevierStyleBold">FACTORES ASOCIADOS AL LUGAR DE FALLECIMIENTO DE LOS PACIENTES CON C&#193;NCER&#46; GRANADA&#44; 1995-99</span></p><p class="elsevierStylePara"> Maria Jos&#233; S&#225;nchez&#42;&#44; Maria Teresa Guerrero&#42;&#44; Elena Corpas&#42;&#44; Carmen Mart&#237;nez&#42;&#44; Rafael G&#225;lvez&#42;&#42;&#44; Nicol&#225;s Olea&#42;&#42;&#42;</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#42;Registro de C&#225;ncer de Granada&#44; Escuela Andaluza de Salud P&#250;blica&#44; Granada&#44; Espa&#241;a&#46; &#42;&#42;Hospital Universitario Virgen de las Nieves&#44; Granada&#44; Espa&#241;a&#46; &#42;&#42;&#42;Hospital Universitario San Cecilio&#44; Granada&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Antecedentes&#58;</span> Diversos estudios han mostrado la preferencia de los pacientes con c&#225;ncer por permanecer en su domicilio hasta la muerte&#46; La experiencia en diversas &#225;reas en pa&#237;ses desarrollados muestra que el porcentaje de los que mueren en su domicilio es inferior al 50&#37;&#46; El lugar de la muerte est&#225; condicionado por factores sociodemogr&#225;ficos&#44; caracter&#237;sticas del propio tumor o duraci&#243;n de la supervivencia&#44; pero tambi&#233;n por la organizaci&#243;n de la atenci&#243;n domiciliaria o de los cuidados paliativos&#46;</p><p class="elsevierStylePara"> Los objetivos fueron&#58; <span class="elsevierStyleItalic">1&#41;</span> conocer el lugar de la muerte &#40;hospital o domicilio&#41; de todas las personas fallecidas en el a&#241;o 1999 y diagnosticadas por primera vez de c&#225;ncer entre 1995 y 1999&#44; residentes en la provincia de Granada&#44; <span class="elsevierStyleItalic">2&#41;</span> identificar los factores asociados al lugar de la defunci&#243;n&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">M&#233;todos&#58;</span> Estudio descriptivo transversal de base poblacional&#46; Se incluyeron todos los fallecidos en el a&#241;o 1999&#44; diagnosticados por primera vez de c&#225;ncer en el per&#237;odo 1995-1999&#44; residentes en la provincia de Granada&#46; La informaci&#243;n se obtuvo del Registro de C&#225;ncer de Granada&#46; Para la codificaci&#243;n de la localizaci&#243;n anat&#243;mica del c&#225;ncer se utiliz&#243; la Clasificaci&#243;n Estad&#237;stica Internacional de Enfermedades y Problemas Relacionados con la Salud &#40;CIE-10&#41;&#46; Las principales variables de estudio fueron&#58; edad&#44; g&#233;nero&#44; lugar de residencia &#40;municipios mayores o menores de 20&#46;000 habitantes&#41;&#44; a&#241;o de diagn&#243;stico&#44; localizaci&#243;n del tumor&#44; tiempo de supervivencia desde el diagn&#243;stico&#44; ingresos hospitalarios durante el a&#241;o del fallecimiento&#44; contacto con Unidad de Cuidados Paliativos y&#47;o del Dolor &#40;UCP&#41;&#46; Para conocer los factores asociados al lugar de la muerte se realiz&#243; un an&#225;lisis de regresi&#243;n log&#237;stica multivariante&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> Se incluyeron 1&#46;393 fallecimientos del a&#241;o 1999&#46; M&#225;s del 50&#37; eran hombres y un 80&#37; eran mayores de 60 a&#241;os&#46; Los c&#225;nceres m&#225;s frecuentes fueron los de pulm&#243;n&#44; colon-recto y pr&#243;stata en hombres y los de mama&#44; colon-recto y leucemias en mujeres&#46; El 38&#37; falleci&#243; en el hospital&#44; si bien en un 10&#37; adicional de los casos el fallecimiento se produjo en un per&#237;odo inferior a 3 d&#237;as desde el alta hospitalaria&#46; Los pacientes con leucemias y linfomas fallecieron con mayor frecuencia en el hospital&#46; En el an&#225;lisis multivariante&#44; los sujetos que presentaron mayor riesgo de morir en el hospital fueron los menores de 60 a&#241;os&#44; residentes en Granada ciudad o grandes n&#250;cleos urbanos&#44; que hab&#237;an tenido un ingreso hospitalario en el a&#241;o 1999&#44; que no hab&#237;an tenido contacto con la UCP y cuya supervivencia era inferior a 2 meses&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> Del total de los pacientes con cancer fallecidos en el a&#241;o 1999&#44; casos incidentes del periodo 1995-1999&#44; un 38&#37; falleci&#243; en el hospital&#46; El lugar de fallecimiento se asoci&#243; con la edad&#44; el &#225;mbito de residencia&#44; la duraci&#243;n de la supervivencia y la asistencia en Unidades de Cuidados Paliativos&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">128</span><span class="elsevierStyleBold">ATTITUDES OF POLISH WOMEN TO SCREENING MAMMOGRAPHY PROGRAMS</span></p><p class="elsevierStylePara"> Ingrid Rozylo-Kalinowska<span class="elsevierStyleSup">1</span>&#44; Pawel Kalinowski<span class="elsevierStyleSup">2</span>&#44; Alina Bochenska<span class="elsevierStyleSup">3</span>&#44; T&#46; Katarzyna Rozylo<span class="elsevierStyleSup">4</span></p><p class="elsevierStylePara"><span class="elsevierStyleItalic"><span class="elsevierStyleSup">1</span>2nd Department of Medical Radiology&#44; University Medical School of Lublin&#44; Lublin&#44; Poland&#46; <span class="elsevierStyleSup">2</span>Department of Epidemiology&#44; University Medical School of Lublin&#44; Lublin&#44; Poland&#46; <span class="elsevierStyleSup">3</span>Private Medical Practice&#44; Wyszk&#243;w&#44; Poland&#46; <span class="elsevierStyleSup"> 4</span>Department of Dental and Maxillofacial Radiology&#44; University Medical School of Lublin&#44; Lublin&#44; Poland&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> Breast carcinoma is one of the most prevalent malignant neoplasms in Polish women&#46; The etiology of the disease has not been fully understood therefore there are no methods of primary prevention&#44; and secondary prevention in the form of screening mammography is used&#46; The success of a screening program depends for the most part on attendance rate of the patients as well as their attitude towards the examination&#46; The objective of the study was to examine the knowledge on purposefulness of mammography among peri-menopausal women as well as to determine their psychological attitude towards planned screening mammography&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> Two-hundred and fifty women inhabitants of Wyszk&#243;w administrative unit in Poland&#44; who attended free screening mammography on the basis of contract with Mazovia Sick Fund in the year 2002&#44; comprised the material&#46; Anonymous questionnaire was filled after an informed consent&#44; directly before the mammographic procedure&#46; The questionnaire contained elements of the Psychological Consequences Questionnaire &#40;PCQ&#41;&#46; The data were analyzed statistically taking into account such demographic data as age&#44; education&#44; marital status&#44; and place of residence&#46; There was analyzed the knowledge of patients on mammography&#44; their motivation for attending such examinations&#44; psychological attitude connected with possible consequences of the result of the procedure&#44; anxiety associated with pain caused by the mammographic technique&#44; as well as declared will to regularly repeat the screening in future&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> The majority of the patients attended screening mammography because it was free as well as because they feared an undetected neoplastic lesion&#46; The women were often preoccupied with the influence of the expected result of mammography on their future&#46; Most of the patients were decided on attending screening mammography in future despite psychological discomfort and physical pain caused by the procedure&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> It was found that psychological attitude influences attendance rate in screening mammography and the decision on further participation in screening programs&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">129</span><span class="elsevierStyleBold">C&#193;NCER EN LA POBLACI&#211;N ANCIANA DE LA REGI&#211;N DE MURCIA</span></p><p class="elsevierStylePara"> Maria Dolores Chirlaque&#44; Carmen Navarro&#44; Miguel Rodr&#237;guez&#44; Jacinta Tortosa&#44; Isabel Valera&#44; Encarnaci&#243;n P&#225;rraga</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Servicio de Epidemiolog&#237;a&#44; Consejer&#237;a de Sanidad y Consumo de la Regi&#243;n de Murcia&#44; Murcia&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introducci&#243;n&#58;</span> La incidencia de c&#225;ncer en la poblaci&#243;n anciana rara vez es estudiada&#44; agrup&#225;ndose en mayores de 75 &#243; 85 a&#241;os&#46; La poblaci&#243;n mayor de 65 a&#241;os en la Regi&#243;n de Murcia registra un incremento del 11&#44;8&#37; al 14&#44;3&#37; en los censos de poblaci&#243;n de 1991 y 2001&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Objetivo&#58;</span> An&#225;lisis detallado de la incidencia de c&#225;ncer en los ancianos de la Regi&#243;n de Murcia&#44; y medici&#243;n de la exhaustividad y exactitud de la informaci&#243;n&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">M&#233;todos&#58;</span> Datos obtenidos del Registro de C&#225;ncer poblacional de la Regi&#243;n de Murcia&#46; Casos incidentes del periodo 1993-96&#46; La clasificaci&#243;n de los tumores utilizada es la CIE-10&#46; Incidencia por grupos de edad &#40;65&#44; 70&#44; 75&#44; 80&#44; 85&#44; 90&#44; 95&#44; 100 o m&#225;s&#41; en ambos sexos de las localizaciones tumorales m&#225;s frecuentes por 100&#46;000 habitantes&#46; La poblaci&#243;n se ha obtenido de una estimaci&#243;n intercensal de los censos de 1991 y 2001&#46; La exhaustividad se mide mediante el porcentaje de casos notificados a partir del certificado de defunci&#243;n&#44; la raz&#243;n mortalidad&#47;incidencia y el porcentaje de casos confirmados microsc&#243;picamente&#46; La exactitud de los datos es medida por el porcentaje de casos cuya &#250;nica fuente es el certificado de defunci&#243;n y casos cuya localizaci&#243;n primaria es desconocida&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Resultados&#58;</span> En hombres&#44; el c&#225;ncer de tr&#225;quea&#44; bronquios y pulm&#243;n presenta la mayor tasa en el grupo de 85-89 a&#241;os con 543 casos&#47;100&#46;000 hab&#44; al igual que el de colon &#40;331&#41;&#46; En el grupo de 90-94 a&#241;os alcanzan su mayor incidencia los tumores de pr&#243;stata &#40;792&#47;100&#46;000&#41;&#44; vejiga &#40;517&#41; y recto &#40;275&#41;&#46; En las mujeres&#44; el c&#225;ncer de cuerpo de &#250;tero presenta su mayor incidencia en el grupo de 60-79 a&#241;os &#40;oscilando de 62 a 73&#47;100&#46;000&#41;&#44; el de mama s&#243;lo supera la tasa de 200&#47;100&#46;000 en las mujeres de 70 a 74 a&#241;os y las localizaciones que presenta la mayor tasa a edades muy avanzadas son las de colon y vejiga&#46; El porcentaje de casos con verificaci&#243;n histol&#243;gica cae de 92&#37; en el grupo de 65-69 a&#241;os&#44; al 17&#37; en hombres y 31&#37; en mujeres en el grupo de 95-99 a&#241;os&#46; El porcentaje de casos notificados a partir del certificado de defunci&#243;n es muy elevado y mayor en hombres de 95-99 a&#241;os &#40;67&#37;&#41; que en mujeres &#40;50&#37;&#41;&#46; La raz&#243;n mortalidad&#47;incidencia aumenta de forma acusada en los grupos de mayor edad&#44; as&#237; como tambi&#233;n empeoran el resto de indicadores de calidad apreci&#225;ndose un agravamiento m&#225;s notable en los hombres en el grupo 80-84 a&#241;os y en las mujeres en el de 85-89&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusiones&#58;</span> En los muy ancianos aumenta de forma acusada la incidencia de las localizaciones tumorales m&#225;s frecuentes&#44; principalmente en hombres&#44; a la vez que los indicadores de calidad muestran que la exhaustividad y la exactitud de los datos es bastante menor&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">130</span><span class="elsevierStyleBold">CHANGE-POINTS IN COHORT AND PERIOD EFFECTS ON MORTALITY TRENDS FROM RENAL CANCER IN EUROPE</span></p><p class="elsevierStylePara"> Napole&#243;n P&#233;rez-Farin&#243;s&#44; Roberto Pastor-Barriuso&#44; Gonzalo L&#243;pez-Abente Ortega</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">Centro Nacional de Epidemiolog&#237;a&#44; Instituto de Salud Carlos III&#44; Madrid&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Background&#58;</span> Recent studies from Europe have shown a global increase in age-adjusted mortality rates from renal cancer&#44; but time trends differ in their shapes among the different European countries&#46; Although age&#44; period&#44; and cohort analyses are useful to graphically display the effect of each individual component&#44; the visual identification of trend changes with this method is subjective&#46; The aim of this study is to formally detect and estimate change-points in cohort and period effects&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> Renal cancer deaths &#40;ICD-9 code 189&#41; and person-years at risk by sex and 5-year age groups were obtained from the WHO database&#46; Data were aggregated in 5-year periods from 1969 to 1999 according to the available data for each country&#46; The 15 European countries with complete series were grouped in 4 homogeneous regions&#58; Nordic &#40;Denmark&#44; Finland&#44; Norway&#44; Sweden&#41;&#44; Central-West &#40;Austria&#44; France&#44; Holland&#44; Ireland&#44; Switzerland&#44; United Kingdom&#41;&#44; East &#40;Bulgaria&#44; Hungary&#41;&#44; and South &#40;Greece&#44; Italy&#44; Spain&#41;&#46; Age- and country-adjusted log-linear Poisson models were fitted within each region to test for the existence of a change-point in cohort and period curvatures&#46; The model&#44; that was implemented in S-Plus&#44; consists of two intersecting linear trends with a smooth transition at an unknown change-point and it provides&#58; &#40;a&#41; the significance level of the test for the change-point&#44; &#40;b&#41; the estimate and 95&#37; CI for the location of the change-point&#44; and &#40;c&#41; the estimates and 95&#37; CIs for annual percentage changes in death rates below and above the estimated change-point&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> Among men&#44; significant changes in cohort and period effects were detected in Nordic&#44; Central-West and South regions &#40;P values &#60; 0&#46;001&#41;&#46; In the Nordic region&#44; the change-point was estimated in 1923 cohort &#40;95&#37; CI 1917 - 1935&#41;&#44; with a 0&#46;37&#37; annual increase below that cohort &#40;0&#46;04 - 0&#46;69&#37;&#41; and a 1&#46;94&#37; annual decrease above it &#40;1&#46;35 - 2&#46;52&#37;&#41;&#46; The annual increase was significantly attenuated from 3&#46;00&#37; below to 0&#46;64&#37; above 1909 cohort for the Central-West region&#44; and from 4&#46;29&#37; below to 0&#46;70&#37; above 1920 cohort for the South region&#46; In the East region&#44; no evidence of change-points in cohort and period effects was detected &#40;P values &#61; 1&#46;00&#41;&#44; with a net annual increase of 2&#46;69&#37; &#40;2&#46;22 - 3&#46;15&#37;&#41;&#46; Within each region&#44; patterns for period and cohort effects were consistent&#46; Time trends for women were fairly similar to those obtained among men&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Discussion&#58;</span> Our results show that&#44; for cohorts born after about 1920&#44; death rates from renal cancer have decreased or levelled off in most European countries&#44; except for those in the eastern region&#44; where the increasing trend continues&#46; Although several factors&#44; such as the development of new diagnostic tests&#44; may contribute to renal-cancer mortality trends&#44; our results suggest that smoking patterns may largely be responsible for the observed differences among European countries&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">131 ADVANCES ON THE STUDY OF SMALL AREA CANCER MORTALITY IN SPAIN</span></p><p class="elsevierStylePara"> Valent&#237;n Hern&#225;ndez&#44; Gonzalo L&#243;pez-Abente&#44; M&#46; Poll&#225;n&#44; N&#46; Aragon&#233;s&#44; B&#46; P&#233;rez-G&#243;mez</p><p class="elsevierStylePara"><span class="elsevierStyleItalic">&#193;rea de Epidemiolog&#237;a Ambiental y C&#225;ncer&#44; Centro Nacional de Epidemiolog&#237;a&#44; Madrid&#44; Espa&#241;a&#46;</span></p><p class="elsevierStylePara"><span class="elsevierStyleBold">Introduction&#58;</span> In geographical studies&#44; the choice of a large-sized administrative spatial unit &#40;such as provinces&#41; tends to dilute the pattern&#44; there being a tendency for differences in risk between smaller intra-provincial areas to be mutually offset&#46; The study of smaller-sized and more homogeneous areas &#40;e&#46;g&#46;&#44; towns&#41; can be useful as a technique for detecting underlying environmental problems&#46; Against this&#44; the choice of town as unit of analysis poses the problem of low numbers of cases&#44; and the use of classic indicators may yield unstable results&#46; The usefulness of smoothed estimators &#40;empirical Bayes&#44; full Bayes&#41; in such a situation has been acknowledged&#46; Our objective is to show the possibility to analyse the mortality pattern of all Spanish towns using a unique regression model&#44; thus obtaining a smoothed map&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Methods&#58;</span> The study covered cancer deaths registered during the period 1989-1998 in Spain&#46; The number of expected deaths was calculated for each of the 8077 towns&#44; with the overall mortality for Spain by age group and sex taken as reference&#46; Person-years were computed on the basis of the 1991 and 1996 census by sex and five-year age groups&#46; In order to obtain a smoothed image of the municipal mortality&#44; we fitted Poisson spatial models&#44; which included two random effects terms&#58; a&#41; municipal contiguity &#40;spatial term&#41;&#59; and b&#41; municipal heterogeneity&#46; These models belong to the so-called conditional autoregressive &#40;CAR&#41; models for disease mapping&#44; initially proposed by Besag&#44; York and Molli&#233;&#46; The models were fitted using Markov Chain Monte Carlo methods with non-informative priors with the WinBugs software&#46; Adjacency of municipal boundaries were employed as criterion of contiguity&#46; Parameter of goodness of fit were calculated and convergence diagnostics were conducted on a sample of towns of different population sizes using tests included in the library CODA for R&#46; Mortality map for emerging tumours &#40;myeloma&#43;non-hodgkin lymphomas&#41; that share some chemical&#47;physical risk factors is shown to illustrate the procedure&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Results&#58;</span> It has been possible to compile and to obtain the posterior distribution of the relative risk from an unique spatial model including the 8077 Spanish towns and the corresponding 47916 adjacencies&#44; investing affordable computing times&#46; The map shows a diffuse pattern with some areas of apparently higher mortality&#46;</p><p class="elsevierStylePara"><span class="elsevierStyleBold">Conclusions&#58;</span> This strategy presents important advantages as&#58; <span class="elsevierStyleItalic">1&#41;</span> their high spatial resolution&#44; which might be useful for environmental surveillance purposes in some cancer locations&#44; <span class="elsevierStyleItalic"> 2&#41;</span> the decrease of edge effect problems&#44; present in atlases bounded to a province or an autonomous region and&#44; <span class="elsevierStyleItalic">3&#41;</span> the efficiency of the method&#46; More research is necessary to solve problems as the excess of zeros and the selection of priors and contiguity criteria&#46;</p>"
    "pdfFichero" => "138v17nSupl.2a13051580pdf001.pdf"
    "tienePdf" => true
    "multimedia" => array:4 [
      0 => array:8 [
        "identificador" => "tbl1"
        "etiqueta" => "Tabla 1"
        "tipo" => "MULTIMEDIATABLA"
        "mostrarFloat" => true
        "mostrarDisplay" => false
        "copyright" => "Elsevier Espa&#241;a"
        "tabla" => array:1 [
          "tablatextoimagen" => array:1 [
            0 => array:1 [
              "tablaImagen" => array:1 [
                0 => array:4 [
                  "imagenFichero" => "138v17nSupl.2-13051580tab01.gif"
                  "imagenAlto" => 132
                  "imagenAncho" => 667
                  "imagenTamanyo" => 6491
                ]
              ]
            ]
          ]
        ]
        "descripcion" => array:1 [
          "en" => "Incidencia de los linfomas no Hodgkin en la RSG&#46; 1994-1999"
        ]
      ]
      1 => array:8 [
        "identificador" => "tbl2"
        "etiqueta" => "Tabla 2"
        "tipo" => "MULTIMEDIATABLA"
        "mostrarFloat" => true
        "mostrarDisplay" => false
        "copyright" => "Elsevier Espa&#241;a"
        "tabla" => array:1 [
          "tablatextoimagen" => array:1 [
            0 => array:1 [
              "tablaImagen" => array:1 [
                0 => array:4 [
                  "imagenFichero" => "138v17nSupl.2-13051580tab02.gif"
                  "imagenAlto" => 153
                  "imagenAncho" => 664
                  "imagenTamanyo" => 9768
                ]
              ]
            ]
          ]
        ]
        "descripcion" => array:1 [
          "en" => "Supervivencia relativa &#40;SR&#41; de los linfomas no Hodgkin a la RSG&#46; 1994-99&#46;"
        ]
      ]
      2 => array:5 [
        "identificador" => "tbl3"
        "tipo" => "MULTIMEDIATABLA"
        "mostrarFloat" => true
        "mostrarDisplay" => false
        "copyright" => "Elsevier Espa&#241;a"
      ]
      3 => array:5 [
        "identificador" => "tbl4"
        "tipo" => "MULTIMEDIATABLA"
        "mostrarFloat" => true
        "mostrarDisplay" => false
        "copyright" => "Elsevier Espa&#241;a"
      ]
    ]
  ]
  "idiomaDefecto" => "en"
  "url" => "/02139111/00000017000000S2/v0_201302051404/13051580/v0_201302051405/en/main.assets"
  "Apartado" => array:4 [
    "identificador" => "792"
    "tipo" => "SECCION"
    "es" => array:2 [
      "titulo" => "Congreso"
      "idiomaDefecto" => true
    ]
    "idiomaDefecto" => "es"
  ]
  "PDF" => "https://static.elsevier.es/multimedia/02139111/00000017000000S2/v0_201302051404/13051580/v0_201302051405/en/138v17nSupl.2a13051580pdf001.pdf?idApp=WGSE&text.app=https://gacetasanitaria.org/"
  "EPUB" => "https://multimedia.elsevier.es/PublicationsMultimediaV1/item/epub/13051580?idApp=WGSE"
]
Compartir
Publique en esta revista
Información de la revista
Vol. 17. Núm. S2.
XXI REUNIÓN CIENTÍFICA DE LA SOCIEDAD ESPAÑOLA DE EPIDEMIOLOGÍA, CONJUNTA CON LA FEDERACIÓN EUROPEA DE EPIDEMIOLOGIA DE LA ASOCIACIÓN INTERNACIONAL DE EPIDEMIOLOGÍA
Páginas 85-90 (octubre 2003)
Respuestas rápidas
Compartir
Compartir
Descargar PDF
Más opciones de artículo
Vol. 17. Núm. S2.
XXI REUNIÓN CIENTÍFICA DE LA SOCIEDAD ESPAÑOLA DE EPIDEMIOLOGÍA, CONJUNTA CON LA FEDERACIÓN EUROPEA DE EPIDEMIOLOGIA DE LA ASOCIACIÓN INTERNACIONAL DE EPIDEMIOLOGÍA
Páginas 85-90 (octubre 2003)
Acceso a texto completo
Pòsters : Cáncer
Cancer
Visitas
954
Este artículo ha recibido
Información del artículo
Texto completo
Descargar PDF
Estadísticas
Tablas (2)
Tabla 1. Incidencia de los linfomas no Hodgkin en la RSG. 1994-1999
Tabla 2. Supervivencia relativa (SR) de los linfomas no Hodgkin a la RSG. 1994-99.
Mostrar másMostrar menos
Texto completo

Jueves 2 de Octubre / Thursday 2, October

17:00:00 a/to 18:00:00

111FACTORES QUE DETERMINAN EL USO DEL PAPANICOLAOU EN MUJERES MEXICANAS

Rosa María Ortiz Espinosa*, Sergio Muñoz Juárez*, Socorro Márquez Maldonado**, Maria de los Angeles Moron Arellaño**

*Coordinación de Investigación, Secretaria de Salud de Hidalgo, Pachuca, México. **Dirección de Regulación Sanitaria, Secretaria de Salud, Pachuca, México.

Antecedentes: El cáncer cervicouterino es la neoplasia más frecuente en mujeres mexicanas y en Latinoamérica. En México existe un programa poblacional desde hace 20 años, no obstante la mortalidad por esta causa continua siendo constante. Existen factores relacionados con la cobertura y accesibilidad del servicio que influyen en la detección y tratamiento oportuno, pero además se encuentran los relacionados con la aceptabilidad, la tendencia e intensidad de uso del Papanicolaou (PAP) por parte de la población, que influyen en la decisión de no hacerse el PAP. En el área rural, las posibilidades que tiene la mujer para decidir sobre su vida sexual y su auto cuidado son, con frecuencia limitadas, la capacidad general para negociar con éxito las necesidades en torno a la salud de su propio organismo son limitadas.

Objetivo: Identificar los factores que determinan el uso del PAP. Material y métodos: Se realizó un diseño transversal analítico y comparativo. Se aplicó una entrevista estructurada a usuarios de 15 a 49 años, de las unidades de primer nivel, seleccionados aleatoriamente. Se utilizó estadística descriptiva, Ji cuadrada y Regresión logística no condicional.

Resultados: Se encontró que el 31,22% de los encuestados desconocían el PAP y su función, de este grupo, el 82.5% nunca se la habían hecho. La mayor proporción de desconocimiento se observo en los hombres en comparación con las mujeres (45% vs 15,49%), el desconocimiento en los usuarios analfabetos fue 33,1% y de 44,3% en los residentes de municipios de mayor marginación. Cuatro de cada diez varones lo desconocen, y en las mujeres dos de cada diez (P = 0,004). El 46% de los usuarios ignoran los factores de riesgo asociados a cacu. En las usuarias que refirieron nunca haberse efectuado un PAP, las principales causas fueron, porque no lo consideran necesario Razón de Momios (RM) crudo de 2,5 con intervalos de confianza (IC) de 95%: 1,3 a 4.8, el personal no me da confianza RM 4,1 IC 95% 1,3 a 12,3. El conocimiento del cáncer cervico uterino (CaCu) y el saber que es curable fue diferente, en las alguna vez usuarias del Pap y aquellas nunca usuarias(P = 0.000), así como el conocimiento de la prevención del cacu (P = 0.00). Las variables asociadas a la demanda del PAP fueron el desconocimiento del CaCu Razón de Momios (RM) 3,6 IC al 95% (1.7 a 7.7), desconocer que se puede evitar RM 1,71 IC (1.19,2.4), desconocimiento que es curable RM 3,36 IC (1.8,6.2) y ser analfabeta RM 4.24 IC (2.1,8.3).

Conclusiones: El conocimiento previo de la neoplasia, y el saber que es curable son factores primordiales que se deben de incluir en las acciones de información, educación y comunicación (IEC) de los programas de prevención y destacar la necesidad de diseñar estrategias donde se difunda los benéficos del PAP.

112SMOKING, ALCOHOL, AND DENTITION IN THE EPIDEMIOLOGY OF ORAL CANCER IN POLAND

Jolanta Lissowska*, Agnieszka Pilarska**, Pawel Pilarski**, Danuta Samolczyk-Wanyura**, Janusz Piekarczyk**, Alicja Bardin-Mikolajczak*, Witold Zatonski*

*Dept. of Cancer Epidemiology and Prevention, Cancer Center & M Sklodowska-Curie Institute of Oncology, Warsaw, Poland. **2 2nd Maxillofacial Surgery Clinic, Medical Academy, Warsaw, Poland.

The role of smoking, drinking, and dental care on the risk of oral and pharyngeal cancer was investigated in a case-control study conducted in Warsaw, Poland. Cases were 122 patients (including 44 females) aged 23-80 years with incident, histologically confirmed cancer of oral cavity and pharynx. Controls were 124 subjects (including 52 females) admitted to the hospital for different non-neoplastic conditions unrelated to tobacco and alcohol consumption, frequency matched to cases by age and sex. Smoking and drinking were strongly associated with an increased risk of oral cancer. Among consumers of both products, risks of oral cancer tended to combine in a multiplicative fashion and were increased more than 14-fold among those who consumed more than 15 cigarettes and 7 or more drinks per day. Cessation of smoking was associated with reduced risk of this cancer. The risks varied by type of cigarettes smoked, being lower among those consuming filtered cigarettes only (OR=1.6) than non-filter (OR=6.5) or mixed (OR=4.2) cigarettes. After adjustment for tobacco smoking and alcohol drinking, poor dentition as reflected by missing teeth, frequency of dental check-ups and frequency of teeth brushing emerged as a strong risk factors. Number of missing teeth and frequency of dental check-ups and frequency of tooth brushing showed increased ORs of 9.8, 11.9 and 3.2 respectively. Denture wearing per se did not affect oral cancer risk. In terms of attributable risk, smoking accounted for 57% of oral cancer cases in Poland, alcohol for 31%. Attributable risks for low frequency of tooth brushing and dental check-ups were 56% and 47% respectively. In conclusion, smoking and drinking cessation are likely to be effective preventive measures against oral cancer. These findings indicate also that poor oral hygiene may be independent risk factor.

113DECLINING MORTALITY RATES FOR NONMELANOMA SKIN CANCERS IN WEST GERMANY, 1968 THROUGH 1999. AN ANALYSIS OF 11.226 NONMELANOMA SKIN CANCER DEATHS

Andreas Stang, Karl-Heinz Jöckel

Epidemiology Unit, Medical Faculaty, University of Essen, Essen, Germany.

Purpose: Since the primary source of data for cancer registries is the inpatient hospital file, routinely collected statistics on nonmelanoma skin cancer (NMSC) are usually incomplete and not comparable with other forms of cancer. We therefore examined time trends of the nonmelanoma skin cancer mortality for the territory of West-Germany including a population of about 66 million people.

Methods: We analysed the nonmelanocytic skin cancer mortality data (1968-1999) from West-Germany including West-Berlin. We calculated age-specific and age-standardized mortality rates (World Standard Population) and used Poisson regression to estimate underlying age, cohort and period effect.

Results: From 1968 (population size of the territory of the Federal State of Germany before the reunification: 60.0 million) through 1999 (population size: 66.9 million), about 11.226 deaths were attributed to NMSC. The NMSC mortality was greater among men than among women throughout the period studied. The estimated percent annual decrease of the age-standardized nonmelanocytic skin cancer mortality rate was -2.3% (95%CI -2.6; -1.9) among men and -3.5% (95%CI: -4.0; -3.1) among women during the period 1968 through 1999. This decline is mainly due to a rate decrease in people aged 80 years or more. The age-specific estimated annual percent changes of the NMSC mortality rates indicate that the greater decline among women is mainly due to a greater rate decrease in women aged 80 years or more. The change in nonmelanoma skin cancer mortality rates was best explained by age-, cohort- and period effects. The age-specific proportions of skin cancer deaths attributed to NMSC declined in people aged 50 year or more from 1968 through 1999. In the early period from 1968 through 1979, about 58% of the male skin cancer deaths and 61% of the female skin cancer deaths in people age 80 years or more were attributed to NSMC. These proportions declined to 33% and 29%, respectively in the latest period from 1990 through 1999.

Conclusions: The nonmelanoma skin cancer mortality in West-Germany showed a continuous decrease from 1968 through 1999. The favourable mortality decline by birth cohort in the most recent birth cohort is an important indicator of a likely decline in mortality over the next years.

114POLYMORPHISMS G691S / S904S OF RET AS GENETIC MODIFIERS IN CANCER PATIENTS FROM FAMILIES WITH MULTIPLE ENDOCRINE NEOPLASIA TYPE 2A

Marina Pollán1, Mercedes Robledo2, Laura Gil3, Arancha Cebrián2, Sergio Ruiz2, Marta Azañedo2, Javier Benítez2, Javier Menárguez4, Jose María Rojas3

1Área de Epidemiología Ambiental y Cáncer, Centro Nacional de Epidemiología del ISCIII., Madrid, Spain. 2Unidad de Genética, Centro Nacional de Investigaciones Oncológicas (CNIO), Madrid, Spain. 3Unidad de Biología Celular, Centro Nacional de Microbiología del ISCIII, Madrid, Spain. 4Unidad de Anatomía Patológica, Hospital Gregorio Marañón, Madrid, Spain.

Introduction: Multiple endocrine neoplasia type 2A (MEN2A) is associated with specific germline missense mutations in the RET proto-oncogene. It is an autosomal dominant trait with high penetrance and variable clinical expression. Medullary thyroid carcinoma is the main clinical feature, but. there are variations, even between members of the same family, regarding the disease onset and its presentation. Our objective was to explore whether two associated RET polymorphisms, G691S and S904S, could have any influence on the clinical form and the age at onset of the disease.

Methods: G691S (exon 11) and S904S (TCC-TCG, exon 15) polymorphisms of RET were analyzed in 198 individuals corresponding to 35 unrelated Spanish MEN2A families (104 patients with oncogenic MEN 2A mutation and 94 healthy relatives) and in a control population of 653 healthy individuals by amplification and sequencing analysis. In all cases, both polymorphisms co-segregated and were considered as a single variable in subsequent analyses. The prevalence of G691S/S904S polymorphisms was compared in MEN2A cases and their healthy relatives using the corrected Pearson's chi-square test allowing for correlation between members of the same family. In the same way, a possible correlation among cases between these polymorphisms and type of clinical presentation was assessed. The relationship between G691S/S904S polymorphisms and age at diagnosis in MEN2A patients was investigated considering "age" as a continuous variable and also as a dichotomous one, taking 20 years as the cut-off. Differences across G691S/S904S groups were quantified using linear regression and logistic regression. In both instances, robust estimators of variance were used, clustered in families. The same analysis was restricted to index cases or probands.

Results: The studied polymorphisms followed Hardy-Weinberg equilibrium in the control population. Among cases, they were not related with the type of clinical presentation, but homozygous were, on average, ten years younger when they were diagnosed (p = 0.037). In fact, homozygous had an 8-fold probability to be diagnosed at an age before 20 (p = 0.010). Obviously, these results could be biased given that the clinical diagnosis for some members of the same family may be conditioned on the time of diagnosis of the corresponding proband. However, when we focused specifically on index cases, the association between age at onset and homozygote G691S / S904S genotype persisted (p < 0.001), and the OR for being diagnosed before 20 was even stronger (OR = 19.3).

Conclusion: These results suggest that the presence of the RET polymorphisms G691S/S904S seems to act as a genetic modifier causing an early appearance of the disease in MEN2A patients. They could be used as markers in asymptomatic children of MEN2A families guiding time of surgical preventive resection.

115GASTRIC CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREA IN SPAIN, 1975-2000

Nuria Aragonés, Gonzalo López-Abente, Marina Pollán, Beatriz Pérez-Gómez, Valentín Hernández, Mario Cárdaba, Berta Suárez, Alicia Estirado

Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Madrid, Spain.

Introduction: Though gastric cancer mortality has been declining during the last decades in Spain, its evolution might have not been uniform across the country. The objective of this analysis is to describe gastric cancer mortality trends by sex and geographical area in Spain.

Material and methods: Both mortality and population data were obtained from the National Statistics Institute. During the study period, three revisions of the International Classification of Diseases were used (8th, 9th and 10th). According to it, 151-code was considered from 1975 to 1998 (8th-9thICD) and C16-code from 1999 to 2000 (10thICD). Individual records broken down by sex, age group, year of death and province of residence were used to compute age-adjusted mortality rates (European standard population) by sex, year, and Autonomous Community. Joinpoint regression analysis was used to detect changes in trends from 1975 to 2000.

Results: In men, Spain had an overall annual reduction of 3% in gastric cancer mortality from 1975 to 2000. Rates decreased a 4.57% per year until 1982, where the speed of this decline was significantly reduced to 2.56%. In women, the same phenomenon was observed: gastric cancer mortality rates decreased 3.84% per year from 1975-2000, although rates descended a 5.95% each year until 1980, where the fall in mortality slowed down to a 3.51%. Among Autonomous Communities, joinpoint regression analysis did not detect significant changes in trends for most of them neither in men nor in women. There are, however, Autonomous Communities where gastric cancer rates for men stopped decreasing or even increased in recent years, as Asturias, Cantabria or Murcia. When studying age groups from 35-64 and 65+, results were quite similar, with annual changes in both sexes around -3.0 and a statistically significant variation in trend around the early 80's in both age groups. This similarity in trends among different age groups points toward a homogeneous cohort effect for successive generations.

Conclusions: Gastric cancer mortality rates fell down in Spain from 1975 to 2000. There is a significant steady decline from 1975 to the early 80's followed by a less accentuated reduction afterwards. This general pattern of decrease could be observed in men and women of different age groups in most geographical areas. There were nevertheless important differences in the magnitude of rates among geographical areas which persisted until recent years, although these differences between Autonomous Communities tended to diminish along time.

116PREVALENCIA DE VPH Y OTROS FACTORES DE RIESGO PARA LESIONES NEOPLÁSICAS PREINVASORAS

Mireia Diaz Sanchis*, Àngela Twose Lázaro*, Jordi Ponce Sebastià**, M. Dolores Martí Cardona**, Silvia de Sanjose Llongueres*, F. Xavier Bosch i José*

*Servicio de Epidemiología y Registro del Cáncer, Institut Català d'Oncologia, Hospitalet de Llobregat, España. **Servicio de Ginecología, Patología Cervical y Colposcopía, Ciudad Sanitaria y Universitaria de Bellvitge, Hospitalet de Llobregat, España.

Introducción: Las clínicas de colposcopía asociadas a programas de cribaje reciben pacientes remitidas por un espectro de diagnósticos citológicos con variabilidad de lectura importante. La determinación del ADN de VPH puede contribuir a clarificar el pronóstico de las lesiones ambiguas (ASCUS) y quizás de las lesiones de bajo grado (LSIL). Los factores de riesgo convencionales de las lesiones preinvasoras tienen un bajo poder discriminatorio y escasa utilidad pronóstica y de control clínico.

Métodos: Se ha realizado un estudio caso-control entre las mujeres remitidas a una Unidad de Patología Cervical y Colposcopía por citología de cribado patológica (ASCUS, LSIL, HSIL). A todas las mujeres se les realizó una entrevista epidemiológica sobre factores de riesgo para neoplasia cervical, una citología y biopsia de verificación y se obtuvo una muestra cervical para la determinación de VPH de alto riego mediante Captura de Híbridos II. En el análisis estadístico se compararon mujeres con lesiones intraepiteliales de bajo y alto grado versus mujeres con lesiones de significado indeterminado mediante modelos de regresión logística politómica.

Resultados: Hasta el momento se han reclutado 375 mujeres con citología de cribado patológica distribuidas en ASCUS 18.1%, lesión intraepitelial de bajo grado (LSIL) 45,1% y lesión intraepitelial de alto grado (HSIL) 36,8%. El determinante mayor de las lesiones de bajo y alto grado versus lesiones ASCUS es la presencia de VPH HR, la prevalencia es de 21,5% para las mujeres con ASCUS, 44,0% para LSIL (ORLSIL = 2,4 (1,2-4,8)) y 81,8% para HSIL (ORHSIL = 15,3 (7,2-32,7)). Se detecta un aumento de riesgo asociado a haber tenido 4 o más compañeros sexuales (ORLSIL = 3,3 (1,2-9,2); ORHSIL = 5,0 (1,7-14,4)) y al consumo de tabaco (ORLSIL = 2,6 (1,3-5,6); ORHSIL = 3,6 (1,6-8,1)). Aparece un efecto protector con 3 o más partos (ORLSIL = 0,2 (0,1-0,8); ORHSIL = 0,2 (0,1-0,9)). Solamente para HSIL, se halla riesgo asociado a haber tenido relaciones sexuales con compañeros casuales (ORHSIL = 3,0 (1,3-6,9)) y un efecto protector con el uso de preservativos (ORHSIL = 0,5 (0,2-0,9)). Ajustando por VPH desaparece el efecto del número de relaciones sexuales y el uso de preservativo, pero se mantienen el consumo de tabaco, el número de partos y para HSIL, las relaciones sexuales con compañeros casuales.

Conclusiones: La prevalencia de ADN de VPH en lesiones citológicas es sensiblemente distinta a la encontrada en estudios realizados en otras poblaciones y que se han establecido como referencia internacional. Por ejemplo, el estudio ALTS* describe ADN de VPH en 50% de los ASCUS y en 80% de las lesiones LSIL, indicando la variabilidad y la especificidad local en la lectura de la citología. Los resultados de referencia en estudios de triaje deben interpretarse en relación a las características de los diagnósticos citológicos locales. La introducción del test de VPH en clínicas de colposcopía puede ayudar a establecer el pronóstico y la conducta terapéutica con mayor precisión que la investigación sobre factores de riesgo de la paciente.

*J Natl Cancer Inst;92:397-402

117ANÁLISIS DE LA VARIACIÓN GEOGRÁFICA DE LA MORTALIDAD POR CÁNCER DE ESTÓMAGO EN GALICIA

Elisa María Molanes, Mª Eugenia Lado

Servicio de información sobre Saúde Pública, Consellería de Sanidade, Santiago de Compostela, España.

Introducción: En los últimos años, han surgido diferentes métodos estadísticos para estimar y suavizar las razones de mortalidad estándar (RME), y poder así realizar comparaciones más fiables del estado de salud de diferentes áreas geográficas. Los principales objetivos de este estudio son: (1) aplicar un "nuevo" método bayesiano no paramétrico de suavizado de tasas y (2) comparar los resultados con los obtenidos al aplicar el modelo clásico de Besag, York y Mollié (modelo de convolución).

Métodos: Para llevar a cabo este estudio se obtuvieron del Registro de Mortalidad de Galicia, los datos de mortalidad por cáncer de estómago (CIE-9 151 y CIE-10 C16) de la población masculina de Galicia en el período 1995-1999. Se utilizaron las poblaciones a 1 de enero para cada año, municipio y grupo quinquenal de edad, y como población estándar se consideró la población gallega en el período de estudio. Con estos datos se calcularon las RME crudas y se estimaron las RME con ambos modelos. Para ajustar el modelo bayesiano no paramétrico se utilizó el software BDCD y para el modelo de convolución el WinBUGS. Para cada método se representaron geográficamente las estimaciones de las RME obtenidas para cada municipio, así como su significación estadística.

Resultados: La distribución espacial de las RME crudas no define claramente ninguna zona de riesgo de mortalidad por cáncer de estómago. En el mapa obtenido con el modelo bayesiano no paramétrico se detectó como zona con mayor riesgo de mortalidad toda la zona occidental de Galicia y la zona centro de la provincia de Lugo. Sin embargo, al estudiar su significación estadística únicamente se mantuvo como área de alto riesgo parte de la zona occidental de Galicia. En lo que se refiere al mapa basado en el modelo de convolución, se detectó como única zona de riesgo la parte occidental de Galicia y, asimismo, sólo una parte de ella se mantuvo como área de riesgo en su mapa de significación.

Conclusiones: De este estudio se concluye que los mapas que representan las RME estimadas son menos ruidosos que el que representa las RME crudas y que ambos modelos coinciden en detectar como zona de mayor riesgo de mortalidad por cáncer de estómago la costa occidental de Galicia. Aunque los resultados obtenidos son muy similares, el modelo bayesiano no paramétrico se muestra más adecuado que el modelo de convolución para la representación geográfica del mapa de riesgos de áreas pequeñas, ya que además de suavizar el valor de las RME en cada área permite detectar discontinuidades en el mapa.

118INCIDENCIA Y SUPERVIVENCIA RELATIVA DE LOS LINFOMAS NO HODGKIN EN GIRONA 1994-1999

Rafael Marcos Gragera*, Àngel Izquierdo Font*, Cristalina Fernández Fidalgo**, Santiago Gardella**, Mª Loreto Vilardell Gil*, Maria Buxó Pujolràs*, Pau Viladiu Quemada*

*Unitat d'Epidemiologia i Registre de Càncer de Girona, Institut Català d'Oncologia de Girona, Girona, Spain. **Servei d'Hematologia, Instiut Català d'Oncologia de Girona, Girona, Spain.

Objetivo: Conocer la incidencia y supervivencia relativa poblacional de los linfomas no Hodgkin (LNH) en Girona.

Material y métodos: A partir de los datos obtenidos por el Registro poblacional de Cáncer de Girona, se analizó la incidencia y supervivencia de los linfomas no Hodgkin en la Región Sanitaria Girona (RSG). La población cubierta por el registro según el censo de 1996 fue de 518.531 habitantes. Se calculan las tasas de incidencia brutas (T.B) y ajustadas (T.Aj) a la población estándar mundial. Para el cálculo de la supervivencia se hizo un seguimiento de los pacientes hasta 12. 1999. Se calculó la supervivencia relativa, tasa entre la supervivencia observada y la esperada, calculada ésta en función de la mortalidad de la población de Girona. Se utilizó el método de Estève.

Resultados:

Conclusiones: En la RSG los linfomas no Hodgkin ocupan el séptimo lugar en orden de frecuencia, tanto hombres como en mujeres. La incidencia de los LNH es mas alta en los hombres. Cuando comparamos con las cifras de incidencia obtenidas en el resto de registros españoles (EUROCIM, ENCR) observamos, en el caso de los hombres, una incidencia estadísticamente superior de los LNH en la RSG. A nivel internacional la incidencia de los LNH en la RSG es situaría a un nivel intermedio - alto. (Cancer Incidence in Five Continents VII, 1997). La SR de los LNH es similar a la que se da en el resto de registros españoles, europeos y americanos (EUCAN, 1997).

119LUNG CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREAS IN SPAIN. 1975-2000

Mario Cárdaba, Nuria Aragonés, Marina Pollán, Beatríz Pérez, Berta Suárez, Alicia Estirado, Valentín Hernández, Gonzalo López-Abente

Epidemiología Ambiental y Cáncer, Instituto de Salud Carlos III, Madrid, España.

Introduction: Lung cancer mortality has been declining since 1990s in Europe. Spain presents a delayed pattern due to its different phase of tobacco epidemic. The aim of this analysis is to describe lung cancer mortality trends by sex, age and geographical area in Spain.

Material and methods: Lung cancer mortality (International Classification of Diseases (ICD)8th:-9th:162; ICD-10th:C34) and population data were obtained from the National Statistics Institute. Individual records broken down by sex, age, year of death and province of residence were used to compute age-adjusted and age-adjusted truncated (35-64) rates (European standard population) by sex, year, and Autonomous Community. Joinpoint regression analysis was used to detect changes in trends between 1975-2000.

Results: Age-adjusted rates for Spanish males showed an annual percent of change(APC) for 1975-2000 of 2.11%. However, joinpoint analysis detected changes in trends in 1988 and 1994 (1975-1988 APC:3.54%; 1988-1994:1.58%; 1994-2000: -0.35%). Truncated rates showed an increase of 2.05% between 1975-200, although rates increased a 3.3% per year until 1990 and reached a plateau since. Among regions, global rates show a positive APC for period 1975-2000 ranging from 0.6% in Cantabria up to 3.52% in Castile-La Mancha. Truncated rates follow a similar homogeneous pattern. Joinpoint analysis detected significant changes in trends in late 1980s or early 1990s, changing from a clear increase to smooth increments, plateaus or even a decline in rates afterwards. Spanish females, for global rates, showed an annual increase of 0.47% between 1975-2000.Nevertheless, joinpoint analysis detected a change in trend in 1990, moving from an annual decrease of -0.71% to an increase of 2.39%. Truncated rates presented an increase of 1.15% for period 1975-2000, but again an acute change in trend turns up in 1990, when APC shifted sharply from -1.43% to 5.38% annually. Among regions, global rates between 1975-2000 seem irregular. Some suggest a rising slope (Balearic Islands:1.53%; Madrid:2.52%; Basque Country:1.87%). Others show a minimal increase(Catalonia: 0.09%; Castile-Leon: 0.24%) or even a decline (Extremadura: -1.75%, La Rioja: -1.24%, Aragon:-0.65%). Truncated rates exhibit a similar pattern though in general increases are greater. Joinpoint analysis detected changes in global and truncated rates. Some regions imitate the nationwide pattern of decrease-increase while others do not. These changes came about in late 1980s or middle 1990s and are sharper in truncated rates. However, Extremadura and Castile-La Mancha present a continuous decline between 1975-2000.

Conclusions: While time trends in lung cancer mortality in men are levelling off since 1990s, in females they are increasing sharply, specially in 35-64 age group, pointing to the beginning of the epidemic phenomenon of lung cancer in women that is affecting to cohorts born after 1940.

120FEMALE BREAST CANCER MORTALITY TRENDS BY GEOGRAPHICAL AREAS IN SPAIN. 1975-2000

Berta Suárez, Nuria Aragonés, Marina Pollán, Beatriz Pérez-Gómez, Mario Cárdaba, Alicia Estirado, Valentín Hernández, Gonzalo López-Abente

Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Madrid.

Introduction: Breast cancer is the commonest malignancy in women in Spain. Though mortality rates have decreased during the last decade, within the country each region has had a different pattern of mortality. The objective of this investigation is to describe female breast cancer mortality trends by geographical area in Spain.

Material and methods: Female breast cancer mortality (International Classification of Diseases (ICD 8th-9th:174; ICD-10th:C50) and population data were obtained from the National Statistics Institute. Individual records broken down by age group, year of death and province of residence were used to compute age-adjusted mortality rates (European Standard population) and age-adjusted truncated rates (35-64 years) by year and by periods of five years for each Autonomous Community(AC) and for Spain as a whole. Joinpoint regression analysis was used to detect statistically significant changes in trends from 1975 to 2000.

Results: In Spain, age-adjusted mortality rates increased until the 1991-1995 period. Joinpoint analysis detected the change in trend in 1991: while the annual increase until this year was 2.41%, rates begun to decline since (-1.75%). Truncated rates had a quite similar evolution, though in this case the annual decrease from 1992 onwards was bigger (-3.33%). This initial increase in rates, followed by a descending trend, was found across the whole country, excepting Cantabria where mortality rates had a continuous light increase. Joinpoint analysis detected statistically significant changes in trend between 1990-95 in most of the Autonomous Communities. Although apparently rates in Aragon, Valencian Community and Basque Country showed a similar trend, changes were not statistically significant. La Rioja had the highest increase, 3.42% per year, and a very fast decline, -6.35%. It is noteworthy the intense decrease observed in Navarra (-8,28%) while in the other ACs descents ranged between -1,04 in Murcia and -3,4% in Catalonia. Also remarkable was Madrid, where two points of change were identified, defining three different periods: rates had a slow increase until 1985 (1.08%), a sharp ascent till 1988 (12.74%) and a slight decline since that year (1.04%). Truncated rates trends among Autonomous Communities had a similar evolution than all-ages groups rates, though with a higher decreasing slope. Every community had a change in trend but the joinpoint analysis only detected significant changes in 10 of them. Murcia was the community with the highest increase before the joinpoint (3.05%), followed up by La Rioja (2.96%) whose decreasing slope afterwards was very high (-9.33%). Madrid had again a different pattern from the other ACs, similar to the one observed for all-ages groups rates.

Conclusions: Breast cancer mortality rates increased in Spain between 1975-1991 and then declined. Screening programs and the improvement in early diagnostic and therapeutic methods might explain this important decline in mortality rates among women.

121NON-HODGKIN LYMPHOMAS MORTALITY IN MADRID

Berta Suárez*, Gonzalo López-Abente*, Consuelo Ibáñez**, Valentín Hernández*, Mario Cárdaba*, Alicia Estirado*, Nuria Aragonés*, Beatriz Pérez-Gómez*, Marina Pollán*

*National Centre for Epidemiology, Carlos III Institute, Madrid, Spain. **Epidemiology Unit, Madrid Regional Health Authority, Madrid, Spain.

Introduction: To study the spatial pattern of home addresses of deaths by non-hodgkin lymphomas in the Madrid Autonomous Community, considering mortality as a spatial point process.

Material and methods: The mortality registry of the Statistics Institute of the Community of Madrid supplied a database containing information from all death certificates with mention to non-hodgkin lymphomas (NHL) during the period 1991-1997as well as a randomly selected sample of 1500 controls, stratified by year of death and sex. All permanent home addresses of cases and controls were georeferencied in UTM-coordinates. Spatial clusters were detected by means of an approach based on the study of Ripley's K functions differences among cases and controls. In order to identify clusters, we obtained a relative risk surface comparing the kernel-smoothed spatial intensity of the process among cases and controls. Its intersection with tolerance bounds from the constant region wide relative risk hypothesis allow to locate the clusters.

Results: 1502 cases of NHL were registered. The case-control comparison showed a possible borderline significant cluster around a distance of 200 meters. The study of the spatial intensity allows to identify NHL clusters in the south area of Torrejón de Ardoz and in two districts of Madrid city. A classical analysis by logistic regression found the following results for the municipal covariate: Torrejón OR=1.96 (95% CI 0.76-1.08) and Madrid OR=1.20 (95% CI 0.99-1.45).

Conclusions: The study of mortality as a spatial point process may be an useful tool to detect patterns that could remain hidden with lattice data analysis. The identified spatial NHL clusters could match partially with the distribution of AIDS rates in the Madrid Region.

122EVOLUCIÓN DE LA MORTALIDAD POR CANCER DE MAMA EN CATALUNYA, 1991-2000

Xavi Puig, Rosa Gispert, Anna Puigdefàbregas

Servei d'Informació i Estudis, Departament de Sanitat i Seguretat Social, Barcelona, España.

Antecedentes: La evolución de la mortalidad por cáncer de mama en años recientes manifiesta una clara inflexión en la tendencia al aumento significativos mostrado en años anteriores. En esta evolución pueden influir tanto la incidencia de este tumor como su supervivencia, aspecto relacionado con la efectividad de la asistencia sanitaria que se presta a estas pacientes. El objetivo del trabajo es analizar si la evolución de la mortalidad por cáncer de mama por grupo de edad es consistente con la tendencia general de esa causa de mortalidad en Catalunya.

Métodos: Se han empleado las defunciones por cáncer de mama (CIM-9:174; CIM-10:C50) del período 1991-2000 del Registro de Mortalidad del Departament de Sanitat de Catalunya, y la población a partir de estimaciones intercensales y postcensales. Los datos se disponen truncados para las mujeres mayores de 34 años, y estratificados por grupos de edad decenales, siendo el último abierto 85 años y más. Para evaluar la tendencia se han ajustado modelos de poisson para cada grupo de edad.

Resultados: En el período 1991-2000 se registraron 10.116 defunciones por cáncer de mama en mujeres mayores de 34 años. El porcentaje global de cambio anual es de -2,7%. Esta evolución no es homogénea por grupos de edad, en los que se observa un claro gradiente, así las mujeres de 35 a 44 años son las que han experimentado una reducción más acusada, del -5,6%, y paulatinamente se modera la tendencia descendiente con la edad, siendo en las mujeres mayores de 84 años de -0,5%.

Conclusiones: La reducción de las tasas de mortalidad por cáncer de mama a diferentes edades ha sido muy importante en la última década, con un manifiesto gradiente relacionado con la edad. La reducción del la mortalidad de manera consistente en todos los grupos de edad suscribe el efecto beneficioso de las intervenciones sanitarias frente a este tumor.

123CHEMICALS AND ELECTROMAGNETIC FIELDS OCCUPATIONAL EXPOSURE AND RISK OF TESTICULAR CANCER AMONG SWEDISH MEN

Alicia Estirado*, Marina Pollán*, Beatriz Pérez-Gómez*, Per Gustavsson**, Nils Plato**, Girgitta Floderus***, Nuria Aragonés*, Montse Alcalde*, Gonzalo López-Abente*

*National Centre for Epidemiology, Carlos III Institute of Health, Madrid, Spain. **Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden. ***Institute of Environmental Epidemiology, Karolinska Institutet, Stockholm, Sweden.

Introduction: Previous studies have reported an association between testicular cancer and some jobs, suggesting that certain occupational exposures could play an etiological role. The aim of this study is the association between seminoma and nonseminoma tumours and the occupational exposure to chemicals and electromagnetic fields (ELMF) among Swedish men.

Material and methods: The base population for this historical cohort comprised all Swedish men recorded in the 1960 census, who were gainfully employed at the time of the 1970 census, and were still alive and over the age of 24 years on January 1, 1971. The follow-up period was 19 years (1971-1989). The Swedish cancer environmental register was used to compute specific rate numerators, and the 1970 census to compute specific rate denominators. Exposure to 13 chemical factors was assessed by linking each combination of occupation and industrial branch to a Swedish job-exposure matrix (JEM), which classifies them as probable, possible and non exposed. Exposure to ELMF was assessed using a Swedish JEM based on the 100 most common jobs among men. The interaction between chemicals and ELMF was done in the subcohort of subjects with information available for both exposures. Relative risks (RRs) adjusted for age, period, geographical area, town-size and occupational sector were computed using log-linear Poisson models. The same analyses were repeated for young people (<40 years).

Results: During follow-up a total of 748 seminomas and 405 nonseminomas were reported. In the general cohort, 39 seminomas were possibly exposed to peak of pesticides, with RR: 1.17 (CI 95%: 0.70 - 1.94). While exposure to solvents for nonseminomas presented a dose response relationship with RR 1.17 for possible exposed and 1.21 for probable exposed, none of them attained statistical significance. Only 8 nonseminomas were possibly exposed to petroleum products: RR: 1.34 (CI 95%: 0.65 - 2.77), and none was probably exposed to this product. In summary, no statistically significant association was found between chemicals or ELMF and testicular cancer in the general cohort or in the subgroup of younger workers. There was not observed an interaction between ELMF and any of the chemicals studied.

Conclusions: Our results did not corroborate the previously reported increased risks for occupational exposure to solvents, oil mixtures, petroleum products, PAH or metals. Nevertheless, exposure misclassification caused by JEM may have biased the RRs towards the null hypothesis. ELMF did not act as a risk factor or as an effect modifier for testicular cancer in this cohort.

124CÁNCER DE MAMA EN LA PROVINCIA DE CÁDIZ: VARIABLES SOCIODEMOGRÁFICAS, CONDUCTAS EN SALUD Y ESTADIO AL DIAGNÓSTICO

María Victoria García-Palacios*, Encarnación Benítez*, Soledad Márquez**, Antonio Escolar*, Estrella Figueroa***

*Servicio de Medicina Preventiva y Salud Pública, Hospital Universitario Puerta del Mar, Cádiz, España. **Escuela Andaluza de Salud Pública, Granada, España. ***Servicio de Medicina Preventiva y Salud Pública, Hospital Universitario de Puerto Real, Cádiz, España.

Objetivos: Describir los factores sociodemográficos, conductas en salud y estadio posquirúrgico TNM de las mujeres diagnosticadas de cáncer de mama.

Metodología: El estudio prospectivo incluyó todos los casos incidentes de tumor maligno de mama con confirmación histopatológica entre el 1 Enero 1999 y 30 Junio 2000 en mujeres residentes en las áreas hospitalarias de Cádiz y Puerto Real. Mediante entrevista por personal entrenado se midieron variables sociodemográficas (edad, estado civil, nivel de educación, municipio de residencia), clínicas (primer síntoma manifestado y número total de síntomas), del sistema sanitario (distrito, cobertura, tipo de consulta ante el primer síntoma),conductas en salud de la mujer (participación en actividades de detección precoz, autoexploración) y tiempo transcurrido hasta el diagnóstico (retraso debido al paciente: tiempo transcurrido desde la aparición del primer síntoma a la visita a un médico y retraso debido al sistema: desde la visita a un médico al diagnóstico). El estadio tumoral en el momento del diagnóstico se midió mediante clasificación TNM. Se realizó análisis de frecuencias, medias, DE y porcentajes para la descripción de los datos.

Resultados: Se estudiaron 179 pacientes (40 asintomáticas diagnosticadas por screening) con una edad media de 56,9 años (rango: 26-90) siendo el 44,4% mujeres < 50 años. La mayor parte (59,2%) no tenía estudios primarios completos. El 60% de las mujeres presentó un solo síntoma, siendo el más frecuente el bulto mamario. Un 46% de las mujeres con edad en criterio de screening no acudió al mismo. En relación a la autoexploración, el 74% no había oído hablar de este término, aunque la mitad de las mujeres la practicaban habitualmente (55,5%). El 60% de las pacientes acudió al médico en el primer mes desde la percepción de su primer síntoma. (media 75,4; DE 184,35; mediana 8,0; rango 0-1106 días). El tratamiento quirúrgico se realizó tras una media de 96,32 días (DE 130,18, mediana 55,0; rango 2-1045) siendo el estadio posquirúrgico más frecuentemente encontrado el IIa y IIb. El 68% presentó un tumor > 2 centímetros, mientras que solamente el 4,1% presentó un estadio con tumor in situ.

Conclusiones: Las mujeres de este estudio tienen una edad similar a la descrita en la literatura, y la mayoría tiene un nivel educativo bajo. La participación en programas de screening fue menor a la considerada como necesaria para reducir la mortalidad por cáncer (70%), por lo que se deben investigar formas de penetración en estos colectivos y otras posibles estrategias de prevención en las mujeres más jóvenes, no incluidas en estos programas y que representaron casi la mitad de los diagnósticos. El tratamiento de estas pacientes sufrió una demora sanitaria considerable siendo la mayoría diagnosticadas con un tumor > 2 centímetros, momento en el que la enfermedad comienza a crecer más rápidamente.

125FACTORES SOCIODEMOGRÁFICOS, SANITARIOS, CONDUCTA EN SALUD Y ESTADIO DIAGNÓSTICO DEL CÁNCER DE MAMA: COMPARACIÓN SEGÚN DOS CLASIFICACIONES

María Victoria García-Palacios*, Encarnación Benítez*, Soledad Márquez**, Antonio Escolar*, Estrella Figueroa***

*Servicio de Medicina Preventiva y Salud Pública, Hospital Universitario Puerta del Mar, Cádiz, España. **Escuela Andaluza de Salud Pública, Granada, España. ***Servicio de Medicina Preventiva y Salud Pública, Hospital Universitario de Puerto Real, Cádiz, España.

Objetivos: Identificar si existe asociación entre los factores socioeconómicos, sanitarios y conductas en salud previas al diagnóstico, y el estadio posquirúrgico TNM del cáncer de mama según dos estadiajes diferentes.

Metodología: El estudio prospectivo incluyó todos los casos incidentes de cáncer de mama con confirmación histopatológica entre el 1 Enero 1999 y 30 Junio 2000 en mujeres residentes en las áreas hospitalarias de Cádiz y Puerto Real. Se realizó entrevista personal midiendo factores de utilización del sistema sanitario, variables clínicas, participación en actividades de detección precoz, autoexploración, retraso debido al paciente y al sistema y otras variables (edad, estado civil, nivel de estudios...). La variable dependiente TNM se reconvirtió en estadios clásicos (I-IV) y en un segundo estadiaje según diagnóstico precoz (precoz: T0-is-T1, N0-N1, M0 y avanzado: T2-T4, N0-N1,M1). Mediante regresión logística se analizó la relación entre el estadio encontrado y las posibles variables predictoras. Se ha evaluado el ajuste de los modelos.

Resultados: Se estudiaron 179 pacientes, siendo los estadios clásicos más frecuentes el IIa (33,1%) y IIb (21,9%); presentando el 68% un diagnóstico no precoz. Se encontró un mayor riesgo de presentar un tamaño tumoral > 2 cm en las mujeres no diagnosticadas por screening (OR = 2,29). De los 2 modelos de regresión finales (excluyéndose 40 mujeres asintomáticas), se relacionó un mayor riesgo de lesiones > 2 cm (diagnóstico avanzado) con: mujeres no casadas (OR = 4,00), no realizar autoexploración (OR=2,55), presentar más de un síntoma (OR = 3,89) y mujeres cuya decisión de consultar al médico no fuese exclusivamente suya (OR = 2,57). Se asoció un mayor retraso en el sistema en los tumores más pequeños (5 meses, frente 3 meses en los de mayor tamaño). El modelo clásico relacionó con un mayor estadio: un menor nivel de estudios (OR=7,8), no realizar autoexploración mamaria (OR = 2,46), tener un síntoma diferente al bulto mamario (OR = 3,73) y cuando la decisión de acudir al médico no la tomaba exclusivamente la paciente (OR = 2,66).

Conclusiones: La clasificación basada en un diagnóstico precoz (T < 2 cm), manifestó la importancia del screening en mujeres asintomáticas y de la consulta precoz ante la presencia de un primer síntoma, para lograr diagnósticos tumorales menores y susceptibles de cirugía conservadora. No obstante, las pacientes con menor tamaño tumoral sufrieron un mayor retraso en el sistema, cuestionando la preparación de dispositivos para un tratamiento definitivo que amenazaría la utilidad de la detección precoz. La conducta en salud de la mujer, como la discutida autoexploración, se relacionó con un tamaño tumoral menor y estadios menos avanzados. El bajo nivel cultural y el presentar un síntoma diferente al bulto mamario aparecen como el principal predictor de una enfermedad avanzada, por lo que debería plantearse una información dirigida a estos grupos que evitara un retraso excesivo por parte del paciente.

126BREAST CANCER RISK FACTORS, ACCORDING TO JOINT ESTROGEN RECEPTOR/PROGESTERONE RECEPTOR STATUS OF TUMOR

Jennifer A. Rusiecki, Theodore R. Holford, Tongzhang Zheng

Department of Epidemiology and Public Health, Yale University, School of Medicine, New Haven, USA.

Introduction: Most of the risk factors identified for breast cancer have been found to have modest relative risks and account for only a portion of breast cancer cases. Characterization of breast tumors on both estrogen receptor (ER) and progesterone receptor (PR) status suggests distinct biological and clinical profiles. Therefore, analyzing breast cancer as one disease may obscure associations with suspected risk factors and lead to studies finding weak associations. Based on these profiles, it has been hypothesized that tumors responsive for both hormones (ER+PR+) are most closely associated with hormone-related risk factors, tumors unresponsive for both hormones (ER-PR-) are less associated with these risk factors, and receptor discordant tumors (ER+PR- and ER-PR+) have intermediate effects.

Methods: In this case-control study we investigated whether the effects of suspected risk factors for breast cancer varied by joint ER/PR status of tumor, by examining age at menarche, age at first full term pregnancy, nulliparity, lifetime lactation, menopausal status, body mass index, ever use of estrogen, alcohol intake, smoking, family history and race, for four tumor subtypes (ER+PR+, ER-PR-, ER+PR- and ER-PR+). For a given risk factor, odds ratios with respect to the common control group were compared using multiple logistic regression, adjusted for all other risk factors simultaneously. We also compared the ER+PR+ case group to the ER-PR- case group, since it has been hypothesized that they represent the two breast cancer subtypes which differ most substantially.

Results: Among 420 cases and 406 controls, the effects of some risk factors varied by joint ER/PR status. Early age at menarche (<12 years) was most strongly associated with ER-PR+ tumors (OR=2.3; 95%CI, 0.7-8.4). Additionally, women with ER+PR+ tumors were 2.2 times more likely to have experienced early menarche than women with ER-PR- tumors (95%CI, 0.8-6.2). Women who had an older age at their first pregnancy (= 30 years) were most likely to have ER+PR- tumors (OR=2.2; 95%CI, 1.1-4.5). Women who reported ever having consumed alcohol were 3.4 times more likely to have ER+PR+ tumors than ER-PR- tumors (95%CI, 1.4-8.4). Family history of breast cancer was most closely associated with ER+PR+ tumors (OR=1.5; 95%CI, 0.9-2.5) and ER+PR- tumors (OR=1.5; 95%CI, 0.8-2.9).

Conclusions: Although we detected several interesting individual effects, there was no clear pattern of association whereby ER+PR+ tumors were most closely associated with hormonally mediated risk factors and ER-PR- tumors were more closely associated with non-hormonally mediated risk factors.

127FACTORES ASOCIADOS AL LUGAR DE FALLECIMIENTO DE LOS PACIENTES CON CÁNCER. GRANADA, 1995-99

Maria José Sánchez*, Maria Teresa Guerrero*, Elena Corpas*, Carmen Martínez*, Rafael Gálvez**, Nicolás Olea***

*Registro de Cáncer de Granada, Escuela Andaluza de Salud Pública, Granada, España. **Hospital Universitario Virgen de las Nieves, Granada, España. ***Hospital Universitario San Cecilio, Granada, España.

Antecedentes: Diversos estudios han mostrado la preferencia de los pacientes con cáncer por permanecer en su domicilio hasta la muerte. La experiencia en diversas áreas en países desarrollados muestra que el porcentaje de los que mueren en su domicilio es inferior al 50%. El lugar de la muerte está condicionado por factores sociodemográficos, características del propio tumor o duración de la supervivencia, pero también por la organización de la atención domiciliaria o de los cuidados paliativos.

Los objetivos fueron: 1) conocer el lugar de la muerte (hospital o domicilio) de todas las personas fallecidas en el año 1999 y diagnosticadas por primera vez de cáncer entre 1995 y 1999, residentes en la provincia de Granada, 2) identificar los factores asociados al lugar de la defunción.

Métodos: Estudio descriptivo transversal de base poblacional. Se incluyeron todos los fallecidos en el año 1999, diagnosticados por primera vez de cáncer en el período 1995-1999, residentes en la provincia de Granada. La información se obtuvo del Registro de Cáncer de Granada. Para la codificación de la localización anatómica del cáncer se utilizó la Clasificación Estadística Internacional de Enfermedades y Problemas Relacionados con la Salud (CIE-10). Las principales variables de estudio fueron: edad, género, lugar de residencia (municipios mayores o menores de 20.000 habitantes), año de diagnóstico, localización del tumor, tiempo de supervivencia desde el diagnóstico, ingresos hospitalarios durante el año del fallecimiento, contacto con Unidad de Cuidados Paliativos y/o del Dolor (UCP). Para conocer los factores asociados al lugar de la muerte se realizó un análisis de regresión logística multivariante.

Resultados: Se incluyeron 1.393 fallecimientos del año 1999. Más del 50% eran hombres y un 80% eran mayores de 60 años. Los cánceres más frecuentes fueron los de pulmón, colon-recto y próstata en hombres y los de mama, colon-recto y leucemias en mujeres. El 38% falleció en el hospital, si bien en un 10% adicional de los casos el fallecimiento se produjo en un período inferior a 3 días desde el alta hospitalaria. Los pacientes con leucemias y linfomas fallecieron con mayor frecuencia en el hospital. En el análisis multivariante, los sujetos que presentaron mayor riesgo de morir en el hospital fueron los menores de 60 años, residentes en Granada ciudad o grandes núcleos urbanos, que habían tenido un ingreso hospitalario en el año 1999, que no habían tenido contacto con la UCP y cuya supervivencia era inferior a 2 meses.

Conclusiones: Del total de los pacientes con cancer fallecidos en el año 1999, casos incidentes del periodo 1995-1999, un 38% falleció en el hospital. El lugar de fallecimiento se asoció con la edad, el ámbito de residencia, la duración de la supervivencia y la asistencia en Unidades de Cuidados Paliativos.

128ATTITUDES OF POLISH WOMEN TO SCREENING MAMMOGRAPHY PROGRAMS

Ingrid Rozylo-Kalinowska1, Pawel Kalinowski2, Alina Bochenska3, T. Katarzyna Rozylo4

12nd Department of Medical Radiology, University Medical School of Lublin, Lublin, Poland. 2Department of Epidemiology, University Medical School of Lublin, Lublin, Poland. 3Private Medical Practice, Wyszków, Poland. 4Department of Dental and Maxillofacial Radiology, University Medical School of Lublin, Lublin, Poland.

Introduction: Breast carcinoma is one of the most prevalent malignant neoplasms in Polish women. The etiology of the disease has not been fully understood therefore there are no methods of primary prevention, and secondary prevention in the form of screening mammography is used. The success of a screening program depends for the most part on attendance rate of the patients as well as their attitude towards the examination. The objective of the study was to examine the knowledge on purposefulness of mammography among peri-menopausal women as well as to determine their psychological attitude towards planned screening mammography.

Methods: Two-hundred and fifty women inhabitants of Wyszków administrative unit in Poland, who attended free screening mammography on the basis of contract with Mazovia Sick Fund in the year 2002, comprised the material. Anonymous questionnaire was filled after an informed consent, directly before the mammographic procedure. The questionnaire contained elements of the Psychological Consequences Questionnaire (PCQ). The data were analyzed statistically taking into account such demographic data as age, education, marital status, and place of residence. There was analyzed the knowledge of patients on mammography, their motivation for attending such examinations, psychological attitude connected with possible consequences of the result of the procedure, anxiety associated with pain caused by the mammographic technique, as well as declared will to regularly repeat the screening in future.

Results: The majority of the patients attended screening mammography because it was free as well as because they feared an undetected neoplastic lesion. The women were often preoccupied with the influence of the expected result of mammography on their future. Most of the patients were decided on attending screening mammography in future despite psychological discomfort and physical pain caused by the procedure.

Conclusions: It was found that psychological attitude influences attendance rate in screening mammography and the decision on further participation in screening programs.

129CÁNCER EN LA POBLACIÓN ANCIANA DE LA REGIÓN DE MURCIA

Maria Dolores Chirlaque, Carmen Navarro, Miguel Rodríguez, Jacinta Tortosa, Isabel Valera, Encarnación Párraga

Servicio de Epidemiología, Consejería de Sanidad y Consumo de la Región de Murcia, Murcia, España.

Introducción: La incidencia de cáncer en la población anciana rara vez es estudiada, agrupándose en mayores de 75 ó 85 años. La población mayor de 65 años en la Región de Murcia registra un incremento del 11,8% al 14,3% en los censos de población de 1991 y 2001.

Objetivo: Análisis detallado de la incidencia de cáncer en los ancianos de la Región de Murcia, y medición de la exhaustividad y exactitud de la información.

Métodos: Datos obtenidos del Registro de Cáncer poblacional de la Región de Murcia. Casos incidentes del periodo 1993-96. La clasificación de los tumores utilizada es la CIE-10. Incidencia por grupos de edad (65, 70, 75, 80, 85, 90, 95, 100 o más) en ambos sexos de las localizaciones tumorales más frecuentes por 100.000 habitantes. La población se ha obtenido de una estimación intercensal de los censos de 1991 y 2001. La exhaustividad se mide mediante el porcentaje de casos notificados a partir del certificado de defunción, la razón mortalidad/incidencia y el porcentaje de casos confirmados microscópicamente. La exactitud de los datos es medida por el porcentaje de casos cuya única fuente es el certificado de defunción y casos cuya localización primaria es desconocida.

Resultados: En hombres, el cáncer de tráquea, bronquios y pulmón presenta la mayor tasa en el grupo de 85-89 años con 543 casos/100.000 hab, al igual que el de colon (331). En el grupo de 90-94 años alcanzan su mayor incidencia los tumores de próstata (792/100.000), vejiga (517) y recto (275). En las mujeres, el cáncer de cuerpo de útero presenta su mayor incidencia en el grupo de 60-79 años (oscilando de 62 a 73/100.000), el de mama sólo supera la tasa de 200/100.000 en las mujeres de 70 a 74 años y las localizaciones que presenta la mayor tasa a edades muy avanzadas son las de colon y vejiga. El porcentaje de casos con verificación histológica cae de 92% en el grupo de 65-69 años, al 17% en hombres y 31% en mujeres en el grupo de 95-99 años. El porcentaje de casos notificados a partir del certificado de defunción es muy elevado y mayor en hombres de 95-99 años (67%) que en mujeres (50%). La razón mortalidad/incidencia aumenta de forma acusada en los grupos de mayor edad, así como también empeoran el resto de indicadores de calidad apreciándose un agravamiento más notable en los hombres en el grupo 80-84 años y en las mujeres en el de 85-89.

Conclusiones: En los muy ancianos aumenta de forma acusada la incidencia de las localizaciones tumorales más frecuentes, principalmente en hombres, a la vez que los indicadores de calidad muestran que la exhaustividad y la exactitud de los datos es bastante menor.

130CHANGE-POINTS IN COHORT AND PERIOD EFFECTS ON MORTALITY TRENDS FROM RENAL CANCER IN EUROPE

Napoleón Pérez-Farinós, Roberto Pastor-Barriuso, Gonzalo López-Abente Ortega

Centro Nacional de Epidemiología, Instituto de Salud Carlos III, Madrid, España.

Background: Recent studies from Europe have shown a global increase in age-adjusted mortality rates from renal cancer, but time trends differ in their shapes among the different European countries. Although age, period, and cohort analyses are useful to graphically display the effect of each individual component, the visual identification of trend changes with this method is subjective. The aim of this study is to formally detect and estimate change-points in cohort and period effects.

Methods: Renal cancer deaths (ICD-9 code 189) and person-years at risk by sex and 5-year age groups were obtained from the WHO database. Data were aggregated in 5-year periods from 1969 to 1999 according to the available data for each country. The 15 European countries with complete series were grouped in 4 homogeneous regions: Nordic (Denmark, Finland, Norway, Sweden), Central-West (Austria, France, Holland, Ireland, Switzerland, United Kingdom), East (Bulgaria, Hungary), and South (Greece, Italy, Spain). Age- and country-adjusted log-linear Poisson models were fitted within each region to test for the existence of a change-point in cohort and period curvatures. The model, that was implemented in S-Plus, consists of two intersecting linear trends with a smooth transition at an unknown change-point and it provides: (a) the significance level of the test for the change-point, (b) the estimate and 95% CI for the location of the change-point, and (c) the estimates and 95% CIs for annual percentage changes in death rates below and above the estimated change-point.

Results: Among men, significant changes in cohort and period effects were detected in Nordic, Central-West and South regions (P values < 0.001). In the Nordic region, the change-point was estimated in 1923 cohort (95% CI 1917 - 1935), with a 0.37% annual increase below that cohort (0.04 - 0.69%) and a 1.94% annual decrease above it (1.35 - 2.52%). The annual increase was significantly attenuated from 3.00% below to 0.64% above 1909 cohort for the Central-West region, and from 4.29% below to 0.70% above 1920 cohort for the South region. In the East region, no evidence of change-points in cohort and period effects was detected (P values = 1.00), with a net annual increase of 2.69% (2.22 - 3.15%). Within each region, patterns for period and cohort effects were consistent. Time trends for women were fairly similar to those obtained among men.

Discussion: Our results show that, for cohorts born after about 1920, death rates from renal cancer have decreased or levelled off in most European countries, except for those in the eastern region, where the increasing trend continues. Although several factors, such as the development of new diagnostic tests, may contribute to renal-cancer mortality trends, our results suggest that smoking patterns may largely be responsible for the observed differences among European countries.

131 ADVANCES ON THE STUDY OF SMALL AREA CANCER MORTALITY IN SPAIN

Valentín Hernández, Gonzalo López-Abente, M. Pollán, N. Aragonés, B. Pérez-Gómez

Área de Epidemiología Ambiental y Cáncer, Centro Nacional de Epidemiología, Madrid, España.

Introduction: In geographical studies, the choice of a large-sized administrative spatial unit (such as provinces) tends to dilute the pattern, there being a tendency for differences in risk between smaller intra-provincial areas to be mutually offset. The study of smaller-sized and more homogeneous areas (e.g., towns) can be useful as a technique for detecting underlying environmental problems. Against this, the choice of town as unit of analysis poses the problem of low numbers of cases, and the use of classic indicators may yield unstable results. The usefulness of smoothed estimators (empirical Bayes, full Bayes) in such a situation has been acknowledged. Our objective is to show the possibility to analyse the mortality pattern of all Spanish towns using a unique regression model, thus obtaining a smoothed map.

Methods: The study covered cancer deaths registered during the period 1989-1998 in Spain. The number of expected deaths was calculated for each of the 8077 towns, with the overall mortality for Spain by age group and sex taken as reference. Person-years were computed on the basis of the 1991 and 1996 census by sex and five-year age groups. In order to obtain a smoothed image of the municipal mortality, we fitted Poisson spatial models, which included two random effects terms: a) municipal contiguity (spatial term); and b) municipal heterogeneity. These models belong to the so-called conditional autoregressive (CAR) models for disease mapping, initially proposed by Besag, York and Mollié. The models were fitted using Markov Chain Monte Carlo methods with non-informative priors with the WinBugs software. Adjacency of municipal boundaries were employed as criterion of contiguity. Parameter of goodness of fit were calculated and convergence diagnostics were conducted on a sample of towns of different population sizes using tests included in the library CODA for R. Mortality map for emerging tumours (myeloma+non-hodgkin lymphomas) that share some chemical/physical risk factors is shown to illustrate the procedure.

Results: It has been possible to compile and to obtain the posterior distribution of the relative risk from an unique spatial model including the 8077 Spanish towns and the corresponding 47916 adjacencies, investing affordable computing times. The map shows a diffuse pattern with some areas of apparently higher mortality.

Conclusions: This strategy presents important advantages as: 1) their high spatial resolution, which might be useful for environmental surveillance purposes in some cancer locations, 2) the decrease of edge effect problems, present in atlases bounded to a province or an autonomous region and, 3) the efficiency of the method. More research is necessary to solve problems as the excess of zeros and the selection of priors and contiguity criteria.

Descargar PDF
Idiomas
Gaceta Sanitaria
Opciones de artículo
Herramientas
es en

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?