Driver dependent factors and the risk of causing a collision for two wheeled motor vehicles.

OBJECTIVE: To assess the effect of driver dependent factors on the risk of causing a collision for two wheeled motor vehicles (TWMVs). DESIGN: Case control study. SETTING: Spain, from 1993 to 2002. SUBJECTS: All drivers of TWMVs involved in the 181 551 collisions between two vehicles recorded in the Spanish registry which did not involve pedestrians, and in which at least one of the vehicles was a TWMV and only one driver had committed a driving infraction. The infractor and non-infractor drivers constituted the case and control groups, respectively. MAIN OUTCOME MEASURES: Logistic regression analyses were used to obtain crude and adjusted odds ratio estimates for each of the driver related factors recorded in the registry (age, sex, nationality, psychophysical factors, and speeding infractions, among others). RESULTS: Inappropriate speed was the variable with the greatest influence on the risk of causing a collision, followed by excessive speed and driving under the influence of alcohol. Younger and older drivers, foreign drivers, and driving without a valid license were also associated with a higher risk of causing a collision. In contrast, helmet use, female sex, and longer time in possession of a driving license were associated with a lower risk. CONCLUSIONS: Although the main driver dependent factors related to the risk of causing a collision for a TWMV were similar to those documented for four wheeled vehicles, several differences in the pattern of associations support the need to study moped and motorcycle crashes separately from crashes involving other types of vehicles.

Risk compensation theory and voluntary helmet use by cyclists in Spain.

OBJECTIVE: To obtain empirical data that might support or refute the existence of a risk compensation mechanism in connection with voluntary helmet use by Spanish cyclists. DESIGN: A retrospective case series. SETTING: Spain, from 1990 to 1999. SUBJECTS: All 22 814 cyclists involved in traffic crashes with victims, recorded in the Spanish Register of Traffic Crashes with Victims, for whom information regarding helmet use was available. MAIN OUTCOME MEASURES: Crude and adjusted odds ratios for the relation between committing a traffic violation and using a helmet. RESULTS: Fifty four percent of the cyclists committed a traffic violation other than a speeding infraction. Committing a traffic violation was associated with a lower frequency of helmet use (adjusted odds ratio (aOR) 0.63, 95% confidence interval (CI) 0.58 to 0.69). Cycling at excessive or dangerous speed, a violation observed in 4.5% of the sample, was not significantly associated with helmet use either alone (aOR 0.95, 95% CI 0.56 to 1.61) or in combination with any other violation (aOR 0.97, 95% CI 0.79 to 1.20). CONCLUSIONS: The results suggest that the subgroup of cyclists with a higher risk of suffering a traffic crash are also those in which the health consequences of the crash will probably be higher. Although the findings do not support the existence of a strong risk compensation mechanism among helmeted cyclists, this possibility cannot be ruled out.

Influence of driver nationality on the risk of causing vehicle collisions in Spain.

STUDY OBJECTIVE: To estimate the association between driver nationality and the risk of causing a collision between vehicles in motion. DESIGN: Retrospective, matched by collision, case-control study. SETTING: Collisions that occurred in Spain during the period from 1990 to 1999 were studied. PARTICIPANTS: Responsible (case) and non-responsible (control) drivers identified in the databases of the Dirección General de Tráfico (General Traffic Directorate) who were involved in a collisions between two or more four wheeled vehicles in motion, in which only one of the drivers had committed a traffic violation. MAIN RESULTS: Crude odds ratios (ORs) for the effect of driver nationality on the risk of causing a collision were significantly higher for foreign drivers than for Spanish drivers, and ranged from a minimum of 1.19 (95% CI 1.09 to 1.29) for Portuguese drivers to a maximum of 2.06 (1.88 to 2.27) for British drivers. Corresponding adjusted ORs were slightly lower, but were still significantly higher than 1 for all nationalities except Italian, Belgian, and American (USA). Adjusted ORs were usually higher for collisions that occurred in urban areas than on open roads. CONCLUSIONS: Authorities responsible for traffic safety, and drivers in general, should consider foreign drivers in Spain at particularly high risk for causing collisions, especially in urban areas.

Application of the induced exposure method to compare risks of traffic crashes among different types of drivers under different environmental conditions.

The authors used the induced exposure method to compare risks of traffic crashes among different types of Spanish drivers under different environmental conditions. The authors analyzed traffic crashes recorded by the Spanish Dirección General de Tráfico for the years 1991 and 1992 to compare proportions of drivers in different age/sex categories who were involved in single-vehicle and multivehicle crashes under different psychological and physical conditions. Crash risk was 1.42- to 2.35-fold greater in men than in women, depending on driver category and environmental factors. Risk was also significantly higher in the 18- to 24-year-old age group (1.75- to 2.87-fold greater than in drivers aged 25-49 years) and under abnormal psychological-physical conditions (1.69- to 4.10-fold greater among drivers under the influence of alcohol). Twilight and night driving, driving in urban areas, and driving on weekends and legal holidays were also associated, though nonsignificantly, with a slightly higher traffic crash risk. These findings are consistent with earlier reports, and they support the usefulness of the induced exposure method as an easy and economical tool with which to analyze data contained in traffic crash records.

Proportion of hospital deaths potentially attributable to nosocomial infection.

OBJECTIVE: To determine the fraction of hospital deaths potentially associated with nosocomial infection (NI). DESIGN: A matched (1:1) case-control study. SETTING: An 800-bed, tertiary-care, teaching hospital. PATIENTS: All patients older than 14 years who were admitted to the hospital between January 1, 1990, and January 1, 1991, were eligible. All 524 consecutive deaths that occurred in the hospital comprised the case group. For each case, a control patient was matched for primary admission diagnosis and admission date. OUTCOME MEASURES: The proportion of hospital deaths potentially associated with NI was estimated from the population attributable risk (PAR) adjusted for age, gender, service, severity of illness, length of stay, and quality of the medical record. RESULTS: For stays longer than 48 hours, the PAR for all NIs was estimated to be 21.3% (95% confidence interval [CI95], 16.8%-30.5%). The greatest proportion of deaths potentially associated with NIs was observed in patients with only one infection (PAR, 15.0%; CI95, 10.9%-22.6%) and bacteremia or sepsis (PAR, 7.7%; CI95, 4.6%-11.6%). CONCLUSIONS: NIs are associated with a large proportion of intrahospital deaths. This information may help clinicians and healthcare managers to assess the impact of programs for the prevention and control of NIs on intrahospital death.

Predictive value of a screen for gestational diabetes mellitus: influence of associated risk factors.

BACKGROUND: There is a need for solid evidence of the relative advantages of universal vs. selective screening for gestational diabetes mellitus. Our study of a broad obstetric population determines the positive predictive value of the 50-g oral glucose challenge test for screening in the presence and absence of classical gestational diabetes risk factors. METHODS: A retrospective cohort study was carried out with a total of 2,574 pregnant women. Clinical information was obtained from hospital records and each patient's medical history, and gestational diabetes risk factors were quantified for each pregnant woman. The positive predictive value of a screen was determined with respect to the number of risk factors. RESULTS: Age 30 or over, family history of diabetes, obesity (BMI > or =27) and previous fetal macrosomia were established as the most frequent risk factors. Just over half (54.2%) of our population presented one or more risk factors. Screening covered 75% of the population, and was positive in 15% of the cases. Diagnosis was confirmed in 64 cases, 57 of them at risk, and seven with no risk factors. We obtained an overall positive predictive value of 21.8% (CI 17.3-27.0). This figure increases with the number of risk factors, from 12% for the women with no risk factors, to 40% for those presenting three or more risk factors. CONCLUSIONS: A selective screening program that takes into account the clinical background and characteristics of each pregnant woman allows a high yield of true positives while reducing the possible undesirable effects associated with false positives.

[Trends in traffic accident mortality in Spain, 1962-1994]. / Evolución de la mortalidad por accidentes de tráfico en España, 1962-1994.

OBJECTIVE: To assess the evolution of the traffic accident mortality rate in Spain from 1962 to 1994, and the role played by its four theoretical components: motorization index (vehicles/population), accidentability index (accidents/vehicles), harmfulness index (victims/accidents) and fatality index (deaths/victims). METHODS: Data from the National Population Census and the Bulletin of the Dirección General de Tráfico were collected to estimate the above mentioned indicators for all accidents and accidents in road and urban zones. Simple and multiple partial correlation coefficients among variables were calculated. Poisson regression models were also obtained. RESULTS: An increasing trend during the whole period was observed for the national traffic accident mortality rate, especially from 1982 to 1989 in the younger age groups, followed by a decrease since 1990. The aforementioned four components were significatively associated with the mortality rate. The strength of this association was especially high for the motorization index and for the harmfulness index when all accidents and road accidents were considered. For urban accidents, the fatality index rate is the component most strongly associated with mortality rate. The role played by the accidentability index in the magnitude of the mortality rate seems less important. CONCLUSIONS: The growing exposure rate to traffic accidents observed in Spain (measured by the motorization index) is not directly influenced by public heath strategies. Therefore, it seems advisable to emphasize the development of measures focused to control the other three components of traffic accident mortality rate, especially those related with harmfulness and fatality.

[Geographical variability in the severity of traffic accidents in Spain]. / Variabilidad geográfica de la gravedad de los accidentes de tráfico en España.

OBJECTIVE: The aim of this study is to analyze the geographical variability of the severity of traffic accidents in Spain, from 1985 to 1994, and to compare several severity indicators. METHODS: The mean values --from 1985 to 1994-- of the following indicators were obtained for each province: mortality index (deaths/accidents), harmfulness rate (victims/accidents), fatality rate (deaths/victims), motorization index (vehicles/inhabitants) and population density. Variability measures among provinces were obtained for each one. Provinces were then grouped in tertiles according to the magnitude of each indicator. Simple correlation coefficients among indicators were calculated. Poisson regression models were obtained, using severity indicators as the dependent variables. RESULTS AND CONCLUSIONS: Geographical variability was especially high for mortality index and fatality rate. For all severity indicators, lower values were found in provinces with the highest population densities and important metropolitan areas. Harmfulness and fatality rates play an independent role upon mortality index.

Evolución de la mortalidad por accidentes de tráfico en España, 1962-1994 / Trends in traffic accident mortality in Spain, 1962-1994

Objetivo: Estudiar la evolución de la tasa de mortalidad por accidentes de tráfico en España desde 1962 hasta 1994 y la capacidad explicativa que sobre ella tienen sus cuatro componentes teóricos: la motorización (vehículos/habitantes), la accidentabilidad (accidentes/vehículos), la lesividad (víctimas/accidentes), y la letalidad (defunciones/víctimas).Métodos: A partir de los Censos de la Población Española y de los Boletines de la Dirección General de Tráfico se han obtenido los datos necesarios para calcular las variables antes referidas, para todos los años del período en estudio, globales, en carretera y en zona urbana. Se han obtenido los coeficientes de correlación de Pearson simple y parcial entre cada indicador y la tasa de mortalidad, y se han aplicado modelos de regresión de Poisson. Resultados: La tasa de mortalidad poblacional muestra una tendencia ascendente durante el período estudiado, y destaca el fuerte incremento de sus valores entre 1982 y 1989, que afectó fundamentalmente a población joven, seguido de un descenso posterior. Los cuatro componentes antes citados se asocian positivamente con la magnitud de la tasa de mortalidad. En el análisis de correlación parcial destaca su asociación con los índices de motorización y de lesividad globales y en carretera, mientras que en zona urbana es la letalidad el factor más estrechamente asociado con ella. El papel de la accidentabilidad parece relativamente menor. Conclusiones: Ya que la tasa de exposición es creciente en nuestro país y escapa a una estrategia de control, es necesario potenciar las medidas que tiendan a disminuir los otros tres componentes de la mortalidad poblacional, específicamente las relacionadas con la lesividad y la letalidad (AU)

Variabilidad geográfica de la gravedad de los accidentes de tráfico en España / Geographical variability in the severity of traffic accidents in Spain

Objetivo: Cuantificar la variabilidad interprovincial de la gravedad de los accidentes de tráfico en España, en el período 1985-1994, y comparar la distribución de diversos indicadores de gravedad. Métodos: Para el período en estudio se han obtenido los valores provinciales medios de tres indicadores de severidad: el índice de mortalidad (defunciones por 1.000 accidentes), la tasa de lesividad (víctimas por 1.000 accidentes) y la tasa de letalidad (defunciones por 1.000 víctimas), así como el índice de motorización (vehículos por 1.000 habitantes) y la densidad de población. Para cada uno se obtuvieron sus correspondientes medidas de dispersión y se clasificaron las 50 provincias en terciles. Se calcularon los coeficientes de correlación entre las cinco variables y se obtuvieron modelos de regresión de Poisson, tomando los indicadores de gravedad como variables dependientes. Resultados y conclusiones: Se observa una importante variabilidad interprovincial en los valores del índice de mortalidad y de la tasa de letalidad. Para todos los indicadores de gravedad, los valores tienden a ser menores en las provincias con elevada densidad de población y grandes concentraciones urbanas. La lesividad y la letalidad aportan componentes explicativos diferentes a la magnitud del índice de mortalidad (AU)

Risk factors in falls among the elderly according to extrinsic and intrinsic precipitating causes.

The aim of this prospective cohort study was to identify the risk factors involved in falls in 190 elderly residents of two geriatric centres in Granada (Andalusia, Spain). Because different types of falls may be associated with different factors, falls were classified according to the precipitating cause, either extrinsic or intrinsic. The incidence density and the ratios for crude and adjusted density were calculated. Cox proportional risk analysis was used to calculate adjusted incidence density ratios. Of the 121 falls identified, 63 (52.1%) had a extrinsic precipitating cause, 43 (35.5%) had an intrinsic precipitating cause, and no precipitating cause was determined in 15 falls. The rate of falls with an extrinsic precipitating cause was 0.39 per person per year, while falls with an intrinsic precipitating cause showed a frequency of 0.27 per person per year. For falls with an extrinsic precipitating cause, the most significant risk factors were: age, diabetes mellitus, a history of falling, and treatment with neuroleptics or oral bronchodilators. The number of illnesses acted as a protective factor. For falls with an intrinsic precipitating cause, the independent risk factors were: age, diabetes, dementia, alterations of gait and balance, previous falls, and treatment with digitalins, neuroleptics or antidepressants. These results suggest that the susceptibility to a fall with an intrinsic precipitating cause is easier to identify and has a greater potential for being controlled.

[Risk factors associated with falls in institutionalized elderly population. A prospective cohort study]. / Factores de riesgo de caídas en una población anciana institucionalizada. Estudio de cohortes prospectivo.

BACKGROUND: Falls in elderly subjects are a frequent cause of morbidity and mortality. The present study investigated risk factors associated with falls in institutionalized elderly subjects. PATIENTS AND METHODS: The subjects of this prospective cohort study were 190 persons aged 65 years or older, who were able to walk unassisted. We recorded sociodemographic variables, morbidity, drugs, functional capacity, cognitive status, gait and balance. Crude and adjusted density ratios (DR) were calculated and Cox proportional risk analysis was performed. RESULTS: A total of 121 falls occurred in 7 persons, for a fall rate of 0.75 per person per year. Independent risk factors identified with multivariate analysis were diabetes mellitus (DR = 3.6), slow gait (DR = 1.7), failed Romberg test (DR = 3.2) and fall(s) during the previous 12 months (DR = 1.9). Drugs that were associated with significant risk were digitalis, antiarrhythmics, neuroleptics, antidepressants and oral bronchodilators. A higher number of chronic diseases acted as a protective factor. CONCLUSIONS: The main risk factors associated with falls in elderly were easy to identify. Analysis of these factors can help to identify elderly persons with a high risk of falls, and thus focus appropriate preventive measures on these subjects.

Life-style factors associated with changes in serum lipids in a follow-up study of cardiovascular risk factors.

We studied some life-style characteristics that may predict changes in total serum cholesterol and cholesterol bound to high density lipoproteins (HDL-cholesterol) in a sample of 980 healthy Spanish employees of both sexes who were followed for three years. All workers participated in a multifactorial program aimed at preventing cardiovascular disease. Linear regression models were fitted with changes in total and HDL-cholesterol as the dependent variables, after eliminating variability due to the influence of basal values. The association between lifestyle factors and lipid changes was controlled for dietary modifications. In the multivariate analysis, decreases in body mass index, and in alcohol consumption were associated with significant reductions in total serum cholesterol. Maintaining sports at post-test or starting to practice them in the interim was also significantly and independently associated with favourable changes in serum cholesterol. Leisure-time exercise (p = 0.002) and giving up smoking (p = 0.06) were each associated with increased HDL-cholesterol.

[Frequency, characteristics and consequences of falls in a cohort of institutionalized elderly patients]. / Frecuencia, características y consecuencias de las caídas en una cohorte de ancianos institucionalizados.

OBJECTIVE: Falls in the elderly are a major problem because of their high morbility and mortality rates and health expenditures. However, there are few studies about this problem in our country. The purpose of the present study was to know the frequency, features and consequences of falling among institutionalized elders. METHODS: We carried out a prospective cohort study using a sample of 190 persons aged 65 years and older living in two nursing homes. All subjects underwent a comprehensive evaluation at the onset of the study. During a mean follow-up period of 310 days, all falls were recorded. RESULTS: There were a total of 121 falls in 72 (37.9%) subjects, twenty five of whom (34.7%) experienced two or more falls. The fall incidence per person-year was 0.75. As compared with males, females had a density ratio (DR) of 2 (1.2-3.2; 95% confidence interval-CI). Falling was more frequent in the bedrooms and living rooms (43.8%). DISCUSSION: This results agree with others studies about the high fall incidence in the elderly and bring out new features about circumstances and consequences of falls.

Risk factors for low birth weight: results from a case-control study in southern Spain.

The main objective of this study is to examine the effect of several variables, including altitude of maternal residence, on delivering a low birth weight (LBW) newborn. A case-control study was done. Two hundred forty cases (single newborn weighing less than 2,500 g) and 374 controls (single newborn weighing more than 2,499 g) were included. Information was gathered from the clinical chart of delivering women, through a personal interview and the Spanish Census Bureau (for altitude). Predictors of LBW were assessed through stepwise logistic regression analysis. Several well-known LBW risk factors were identified: hypertension, weight gain during pregnancy, body size (mainly maternal prepregnancy weight), low social class, primiparity, and several conditions (spontaneous delivery, abruptio placentae). Altitude was an independent predictor of LBW at term (more than 37 weeks of gestational age) but not for preterm LBW. Nevertheless, a relationship between altitude and birth weight was not found in controls, although a moderate decreasing gradient with altitude was observed. The limitations of these findings are discussed.

Changes in serum cholesterol in employees after three years of multifactorial intervention.

BACKGROUND: To assess long-term effectiveness of a multifactorial intervention at the work-site on serum cholesterol levels. METHODS: Individualized face-to-face counseling was given to 1,555 employees (76.7% male; mean age = 42.3 years) by occupational physicians at four work-sites. After 3 years, a blinded assessment of the adequacy of the intervention was done. Implementation of the intended intervention by physicians was assessed as adequate in two work-sites (927 employees) and inadequate in the other two (628 employees). Observed changes in serum cholesterol were analyzed in the followed-up individuals. Follow-up rates at each work-site were 78.6% and 44.5% for the adequate intervention, and 85.5% and 60.4% for the inadequate intervention. Changes in serum cholesterol were controlled for potential confounding factors (pre-test levels of risk factors, age, sex, body mass index, educational level, marital status, physical activity and alcohol consumption) by multiple linear regression procedures. RESULTS: When the intervention was adequately performed, serum cholesterol was significantly lowered with a mean reduction of 14.3 mg/dl (95% C.I.: 11.0 to 17.6) in those employees with baseline levels > or = 200 mg/dl. CONCLUSIONS: Adequacy of implementation of work-site programs determines their long-term effectiveness in reducing mean serum cholesterol levels.

Unplanned pregnancy as a major determinant in inadequate use of prenatal care.

BACKGROUND: Previous reports have stressed the importance of social class and education in prenatal care use. Unplanned pregnancy as a determinant of prenatal care use has been insufficiently studied. The objective of this report was to assess whether unplanned pregnancy is an independent predictor of inadequate use of prenatal care. METHODS: A 5% sample of women delivering at a hospital (409 women in the study population) was selected. Data on pregnancy were obtained by personal interview and from clinical charts. Prenatal care was considered inadequate according to the Kessner index. Relative risk (RR) and 95% confidence intervals (CI) were estimated. Stepwise logistic regression analysis was applied to select the independent predictors of inadequate prenatal care use. RESULTS: Prenatal care use was inadequate among 16.4% of the women. Pregnancy was unplanned among 42.8% of the women. Twenty-two percent of women with an unplanned pregnancy used prenatal care inadequately, while 12% of those with planned pregnancies used prenatal care inadequately (RR = 1.9, 95% CI = 1.2-2.9). In crude analysis, inadequate prenatal care use was also related to lower social class, lower education level, no employment outside the home, and multiparity. After adjustment was made for other predictors that were included in a stepwise logistic regression model (maternal education, social class, maternal occupation, parity, and pregnancy-induced hypertension), unplanned pregnancy was a significant risk factor for inadequate use of prenatal care (odds ratio = 2.1, 95% CI = 1.2-3.7) and it was an independent predictor for a delayed first prenatal care visit and for a reduced number of visits. CONCLUSIONS: The results suggest that unplanned pregnancy is a major determinant for inadequate use of prenatal care.

[Comparison of 2 indexes of prenatal care and risk of preterm delivery]. / Comparación de dos índices de asistencia a la atención prenatal y riesgo de parto pretérmino.

OBJECTIVES: A new index the adequacy of prenatal care utilization (APNCU) index, has been proposed to provide a more accurate and comprehensive measure of antenatal care use than the widely used Kessner index. To better understand the value of the two above mentioned indexes as predictors of preterm delivery, we examined their ability to predict women who will or will not deliver before 37 weeks of gestation. METHODS: A case-control study was performed, including 207 cases and 381 controls. Prenatal care was assessed on the basis of the two above mentioned indexes, both taking into account the number of prenatal care visits, the date of the first visit and gestational age. Multiple-factor adjusted odds ratios and their 95% confidence intervals were estimated using logistic regression methods. RESULTS: The Kessner index showed a lineal trend with both crude and adjusted for (the APNCU index) estimates of preterm delivery risk. The APNCU index did not show any linear trend with adjusted for (the Kessner index) estimates of preterm delivery risk. To assess whether the Kessner index added explanatory information to the APNCU index (or vice versa), the APNCU index was regressed on the Kessner index (and viceversa), and a set of residuals was computed for both indexes. In logistic regression analyses, the residuals of Kessner added meaningful information to the APNCU index, whereas the residuals of the APNCU index did not add any relevant information to the Kessner index. There results remained unchanged after controlling for several confounders. CONCLUSIONS: A variation in the definition of adequate prenatal care use changes the association between prenatal care and preterm delivery. The Kessner index showed a better ability for discriminating and predicting the risk of preterm delivery.

[Morbidity, mortality and the potential years of life lost attributable to tobacco]. / Morbilidad, mortalidad y años potenciales de vida perdidos atribuibles al tabaco.

BACKGROUND: The aim of this study was to evaluate the morbidity and mortality attributable to smoking. MATERIAL AND METHODS: A transversal study was performed in a reference population of residents of the province of Granada (Spain). The number of deaths in 1985, the number of patients admitted to public hospitals in 1987 and a sample of out patients in 1989 were gathered from the National Institute of Statistics and hospital registries, and a prospective collection from out patient clinics was also carried out. The diseases studied included: oropharyngeal, esophageal, pancreatic, laryngeal, lung, bladder and kidney carcinomas, ischemic cardiopathy, cerebrovascular stroke, peripheral arteriopathy and chronic obstructive lung disease (COPD). Statistical analysis was performed by populational etiologic fraction and the chi square test. RESULTS: Four point five percent of the total number of annual hospital admissions, 6.3% of the out patients, 15.9% of the mortality and 12.0% of the total number of potential years of life lost (TPYLL) were estimated to be attributable to smoking in individuals over the age of 15 years in the province of Granada. The most frequent causes of morbidity attributable to smoking were COPD and ischemic cardiopathy, and with regards to mortality, ischemic cardiopathy (28.5%), COPD (21.6%) and lung cancer (17.5%). Higher percentages were observed in males than in females and in the group from 46 to 65 years of age. CONCLUSIONS: Smoking is not only responsible for an important percentage of deaths, but also leads to premature death and a reduction in the quality of life which is translated into an excess of health care costs. Since the impact of smoking becomes particularly evident from 46 to 65 years of age, smoking prevention should be centered on the early years of life.

Proportion of hospital deaths associated with adverse events.

OBJECTIVES: To determine the fraction of hospital deaths potentially associated with the occurrence of adverse events (AE). DESIGN: A paired (1:1) case-control study. SETTING: An 800-bed, teaching tertiary care hospital. PATIENTS: All patients older than 14 years admitted to the hospital between January 1, 1990, and January 1, 1991, were eligible. All 524 consecutive deaths (death rate of 3.74%) that occurred in the hospital comprised the case group. For each case, a control patient was matched for both primary diagnosis on admission and admission date. MEASUREMENTS: The proportion of hospital deaths associated with adverse events (defined as problems of any nature and seriousness faced by the patient during hospitalization, and potentially traceable to clinical or administrative management) was estimated from attributable risks adjusted for age, sex, service, severity of illness, length of stay, and quality of the medical record. RESULTS: For stays longer than 48 hours, the adjusted attributable risk for all adverse events was estimated to be 0.51 (0.40-0.61). When the data were stratified according to the category of adverse event, the attributable risks remained significant except for administrative problems. The greatest proportion of deaths associated with adverse events was observed for surgical adverse events [0.56 (0.38-0.71)] and nosocomial infection [0.22 (0.14-0.28)]. CONCLUSIONS: A significant proportion of intrahospital deaths were associated with AE. These results suggest the need to consider programs focused on the prevention of mortality from AE.