The excess mortality related to cardiovascular diseases and cancer among adults pharmacologically treated for diabetes--the 2001-2006 ENTRED cohort.

AIMS: To compare the 5-year mortality (overall and cause-specific) of a cohort of adults pharmacologically treated for diabetes with that of the rest of the French adult population. METHODS: In 2001, 10 000 adults treated for diabetes were randomly selected from the major French National Health Insurance System database. Vital status and causes of death were successfully extracted from the national registry for 9101 persons. We computed standardized mortality ratios. RESULTS: Over 5 years, 1388 adults pharmacologically treated for diabetes died (15% of the cohort, 32.4/1000 person-years). An excess mortality, which decreased with age, was found for both genders [standardized mortality ratio 1.45 (1.37-1.52)]. Excess mortality was related to: hypertensive disease [2.90 (2.50-3.33)], ischaemic heart disease [2.19 (1.93-2.48)], cerebrovascular disease [1.76 (1.52-2.03)], renal failure [2.14 (1.77-2.56)], hepatic failure [2.17 (1.52-3.00)] in both genders and septicaemia among men [1.56 (1.15-2.09)]. An association was also found with cancer-related mortality: liver cancer in men [3.00 (2.10-4.15)]; pancreatic cancer in women [3.22 (1.94-5.03)]; colon/rectum cancer in both genders [1.66 (1.28-2.12)]. Excess mortality was not observed for breast, lung or stomach cancers. CONCLUSIONS: Adults pharmacologically treated for diabetes had a 45% increased risk of mortality at 5 years, mostly related to cardiovascular complications, emphasizing the need for further prevention. The increased risk of mortality from cancer raises questions about the relationship between cancer and diabetes and prompts the need for improved cancer screening in people with diabetes.

Measuring social inequalities in cause-specific mortality in France: comparison between linked and unlinked approaches.

BACKGROUND: Monitoring the time trends in socioeconomic inequalities in mortality by cause is a key public health issue. The aim of this study was to compare methods to measure social inequalities in cause-specific mortality in the French population aged 25-55 years. More specifically, it compares bias and precision related to the use of occupational class declared at the last census (linked data) to the one declared at the time of death on the death certificate (unlinked data). METHODS: We used a representative sample of 1% of the French population. Causes of death were obtained by direct linkage with the French national death registry. Occupational class was classified into eight categories. Taking professionals and managers as the reference, relative risks of mortality by cause and their 95% confidence intervals were estimated using Poisson models for the 1983-1989, 1991-1997, and 2000-2006 periods. The relative risks were calculated with both linked data and exhaustive unlinked data. RESULTS: Over the 2000-2006 period, occupational classes declared at census and on the death certificate were consistent for half of the deaths. Relative risks for manual workers were found to be similar between the two approaches over the 1983-1989 and 1991-1997 periods, and higher for the unlinked approach over the 2000-2006 period. Over the latter period, the order and magnitude of relative risks varied similarly by occupational class and cause of death for both approaches. Confidence intervals obtained from linked data were wide. CONCLUSION: Occupational class derived from the death certificate must be used with caution as a measure for epidemiological purposes and the available linked data do not allow accurate estimates of social inequalities in cause-specific mortality. Other solutions should be considered in order to improve the follow-up of social inequalities in mortality. This would require the collection of educational level on the death certificate or the linkage of the cause of death database with other exhaustive and informative databases.

Deaths associated with acquired haemophilia in France from 2000 to 2009: multiple cause analysis for best care strategies.

Deaths occurring in the context of acquired haemophilia (AH) may be related to inter-connected causes and mechanisms including bleeding, specific or older patient co-morbidities or iatrogenic complications. However, their magnitude remains unknown. This study aimed to determine the respective weight and frequency of the various causes of death in AH. Multiple-cause analysis based on death certificates data is used in this purpose. Over a 10-year period (2000-2009), 121 deaths with AH as a cause were registered in France. All the deaths were of adults (extremes: 47 and 99 years; mean age: 80.7 years). The average number of causes per death certificate was 4.7. AH was the underlying cause of death (UCD) in 69.4% of the cases, and was more frequent in the older subjects. In contrast, before age of 75 years, AH was more often a contributing cause of death. No postpartum or obvious thromboembolism-related deaths were registered. Haemorrhagic shock was the most frequent direct cause of death (DCD), followed by infectious events, cardiac dysfunction, metabolic and nutritional disorders with muscle wasting and decubitus complications, and cancers (52.9%, 26.4%, 7.5%, 5.8% and 4.1%, respectively). However, when AH was not reported as an UCD, infections become the first DCD (32.4%) followed by bleeding events (16.2%). Best prophylactic and curative strategies for infections are particularly required to improve the prognosis in AH. Moreover, as several of its DCD correspond also to steroids side effects, best tolerated immunosuppressant regimen with steroid-sparing agents adjoining are particularly awaited in AH population.

[Knowledge of causes of death in Algeria. Methods and first results]. / Connaissance des causes de décès en Algérie. Etude des décès enregistrés par l'INSP. Méthodes et premiers résultats.

BACKGROUND: This article describes the gradual establishment between 1995 and 2009 of the collection of medical causes of death in Algeria by the National Institute of Public Health (INSP). METHODS: The registration of these causes is based on the WHO certificate's model. The codes and rules of the International Classification of Diseases (ICD10) were used for coding. RESULTS: Initial results for 2007 show that causes of death have been gathered by INSP for 36.5% of the deaths registered by the civil registrar. Among these causes, cardiovascular diseases occupy the first place. Distributions are different by gender and age. CONCLUSION: This study is a first step towards the knowledge of the causes of death in Algeria.

Liver-related deaths in HIV-infected patients between 1995 and 2005 in the French GERMIVIC Joint Study Group Network (Mortavic 2005 study in collaboration with the Mortalité 2005 survey, ANRS EN19).

BACKGROUND: More than 10 years after the introduction of combination antiretroviral therapy (cART), we examined the trend in the proportion of deaths caused by end-stage liver disease (ESLD) in HIV-infected adults in France between 1995 and 2005. DESIGN AND METHODS: In 2005, 34 departments prospectively recorded all deaths in HIV-infected patients who were followed in those departments (around 24 000). RESULTS: were compared with those of four previous cross-sectional surveys conducted since 1995 using the same methodology. Results Among 287 reported deaths in 2005, 100 (35%) were related to AIDS, and 48 (17%) to ESLD. Three out of four patients who died from ESLD-related causes had chronic hepatitis C. Excessive alcohol consumption was reported in approximately half of the patients (48%). At death, 62% of patients had undetectable HIV viral load and the median CD4 count was 237 cells/microL. From 1995 to 2005, the proportion of deaths caused by ESLD increased from 2 to 17% (P<0.001). The proportion of deaths caused by hepatocellular carcinoma increased from 5% in 1995 to 25% in 2005 (P=0.0337). CONCLUSIONS: Over the 10 years from 1995 to 2005, the proportion of deaths caused by hepatitis C virus-related ESLD has increased in HIV-infected patients. ESLD is currently a leading cause of death in this population, with hepatocellular carcinoma representing a quarter of liver-related deaths. Recommendations for the detection of hepatocellular carcinoma should be strictly applied in these patients.

Mortality related to anaesthesia in France: analysis of deaths related to airway complications.

Death certificates from the French national mortality database for the calendar year 1999 were reviewed to analyse cases in which airway complications had contributed to peri-operative death. Respiratory deaths (and comas) found in a previous national 1978-82 French survey (1:7960; 95% CI 1:12,700 to 1:5400) were compared with the death rate found in the present one: 1:48,200 (95% CI 1:140,000 to 1:27,500). In 1999, deaths associated with failure of the breathing circuit and equipment were no longer encountered and no death was found to be related to undetected hypoxia in the recovery unit. Deaths related to difficult intubation also occurred at a lower rate than in the previous report (1:46,000; 95% CI 1:386,000 to 1:13,000) in 1978-82 vs 1:176,000 (95% CI 1:714,000 to 1:46,000) in 1999, a fourfold reduction. In most cases, there were both inadequate practice and systems failure (inappropriate communication between staff, inadequate supervision, poor organisation). This large French survey shows that deaths associated with respiratory complications during anaesthesia have been strikingly reduced during this 15-year period.

Cancer incidence and mortality in France over the period 1980-2005.

BACKGROUND: The objective of this study was to provide updated estimates of national trends in cancer incidence and mortality for France for 1980-2005. METHODS: Twenty-five cancer sites were analysed. Incidence data over the 1975-2003 period were collected from 17 registries working at the department level, covering 16% of the French population. Mortality data for 1975-2004 were provided by the Inserm. National incidence estimates were based on the use of mortality as a correlate of incidence, mortality being available at both department and national levels. Observed incidence and mortality data were modelled using an age-cohort approach, including an interaction term. Short-term predictions from that model gave estimates of new cancer cases and cancer deaths in 2005 for France. RESULTS: The number of new cancer cases in 2005 was approximately 320,000. This corresponds to an 89% increase since 1980. Demographic changes were responsible for almost half of that increase. The remainder was largely explained by increases in prostate cancer incidence in men and breast cancer incidence in women. The relative increase in the world age-standardised incidence rate was 39%. The number of deaths from cancer increased from 130,000 to 146,000. This 13% increase was much lower than anticipated on the basis of demographic changes (37%). The relative decrease in the age-standardised mortality rate was 22%. This decrease was steeper over the 2000-2005 period in both men and women. Alcohol-related cancer incidence and mortality continued to decrease in men. The increasing trend of lung cancer incidence and mortality among women continued; this cancer was the second cause of cancer death among women. Breast cancer incidence increased regularly, whereas mortality has decreased slowly since the end of the 1990s. CONCLUSION: This study confirmed the divergence of cancer incidence and mortality trends in France over the 1980-2005 period. This divergence can be explained by the combined effects of a decrease in the incidence of the most aggressive cancers and an increase in the incidence of less aggressive cancers, partly due to changes in medical practices leading to earlier diagnoses.

Patterns of mortality and their changes in France (1968-99): insights into the structure of diseases leading to death and epidemiological transition in an industrialised country.

BACKGROUND: Epidemiological transition theory is based on a succession of specific "patterns" of causes of death in human societies. However, the reality and consistency of patterns of causes of death in a population at a given moment has never been formally and statistically evaluated. METHODS: Correlation analyses and principal component analysis were used to explore the correlation between age and sex cause-specific death rates and to identify consistent patterns of mortality in France for two periods: 1968-79 and 1988-99. RESULTS: Cause-specific death rates in France from 1988 to 1999 were found to be strongly and consistently correlated across space and time. The analysis outlines four specific patterns: mortality of 45 - 84-year olds, mostly by neoplasms, cardiovascular and digestive diseases; mortality of the oldest old (>84 years); mortality of 25 - 64-year-old men, notably by HIV infection; and mortality by injury and poisoning of 15 - 44-year olds. These patterns, which cover 96% of the total mortality during the period, differ from those for the period 1968-79 when respiratory diseases and conditions affecting children aged <1 year shaped mortality. They also differ substantially from those predicted by classical epidemiological transition theory. CONCLUSION: This study provides evidence for an evolutionary structure of patterns of mortality in contemporary France and therefore suggests using the concept of epidemiological transition in a less simplistic way than is commonly the case. It also shows much stronger interrelationships between diseases leading to death than is usually believed and suggests that current categorisations of cause-specific mortality in populations need reconsideration.

[Coronary mortality in France according to data sources]. / Mortalité coronaire en France selon les sources d'information.

BACKGROUND: Since 1985, two sources of information currently yield coronary disease frequency indicators among the French population: the national cause of death statistics set up by the CépiDC (INSERM), on the one hand, and three registries recording myocardial infarction and coronary deaths as defined by the WHO MONICA Project in three regions (Bas-Rhin, Communauté Urbaine de Lille, Haute-Garonne) on the other hand. Particularly, an inquiry for each possibly coronary death allows the registries to conclude positively (with or without a myocardial infarction), negatively or that no conclusion can be drawn because of insufficient data. The aim of the present work is to analyze concordance between coronary deaths issuing from the two sources according to their definition, while taking into account, or not, multiple causes listed on the death certificates. MATERIAL: and methods: In total, 4,664 deaths occurring in 2000 in the 35-64 year-old population of the three regions identified by the CépiDc were paired with the 812 deaths analyzed by the registries. The MONICA classification was compared with that of the CépiDC which used the ICD 10th Revision of the initial cause or after taking into account multiple causes. In each case, the concordance between the final classifications (coronary deaths or not) and the mortality ratio obtained from the two sources were computed. RESULTS: and conclusions: Eight hundred and six deaths could be paired: 310 with a coronary cause according to the registries, 420 of presumed coronary cause but with insufficient data and 76 of non coronary origin. Whereas the total number of coronary deaths was similar for the two sources, their concordance was relatively low (kappa=0.61). However, when the deaths with insufficient data were included in the MONICA definition, concordance decreased and a large underestimation (59%) of the coronary mortality is given by the national statistics as compared to the registries. Taking into account multiple causes of death and not only the initial cause permitted partly to reduce this underestimation (42%) and to increase concordance (kappa from 0.46 to 0.51). These findings have important consequences for international comparisons concerning coronary disease. Indeed, the MONICA Project showed that the frequency of deaths with insufficient data was especially elevated in France leading to an underestimation of the coronary death rates provided by the national statistics in comparison with other countries, particularly in Europe.

Ecological association between indoor radon concentration and childhood leukaemia incidence in France, 1990-1998.

The objective of this study was to evaluate the ecological association between indoor radon concentration and acute leukaemia incidence among children under 15 years of age in the 348 geographical units (zones d'emploi, ZE) of France between 1990 and 1998. During that period, 4015 cases were registered by the French National Registry of Childhood Leukaemia and Lymphoma. Exposure assessment was based on a campaign of 13 240 measurements covering the whole country. The arithmetic mean radon concentration was 85 Bq/m (range, 15-387 Bq/m) and the geometric mean, 59 Bq/m (range: 13-228 Bq/m). A positive ecological association, on the borderline of statistical significance (P=0.053), was observed between indoor radon concentration and childhood leukaemia incidence. The association was highly significant for acute myeloid leukaemia (AML) (P=0.004) but not for acute lymphocytic leukaemia (ALL) (P=0.49). The standardized incidence ratio (SIR) increased by 7, 3 and 24% for all acute leukaemia, ALL and AML, respectively, when radon concentration increased by 100 Bq/m. In conclusion, the present ecological study supports the hypothesis of a moderate association between indoor radon concentration and childhood acute myeloid leukaemia. It is consistent with most previous ecological studies. Since the association is moderate, this result does not appear inconsistent with the five published case-control studies, most of which found no significant association.

[Neonatal mortality in France: usefulness of a neonatal death certificate]. / La mortalité néonatale en France: bilan et apport du certificat de décès néonatal.

UNLABELLED: A neonatal death certificate was introduced in France in 1997. It provides detailed data on the causes of death and the characteristics of newborn, birth and parents. Our aim was to describe the new results of this certificate. METHOD: All deaths in 1999 in the first 27 days of life were included (N=2036). Certificates were analysed using the usual process, especially following the International Classification of Diseases. RESULTS: The neonatal death certificate was used for 87% of deaths. The proportion of documented items was 96% for gestational age and birthweight, 87% for maternal age and parity and 70% for maternal occupation. Almost three quarters of the deaths occurred in the first 6 days (36.9% in the first 24 hours and 35.1% between one and six days). 30.5% of the died infants were born before 27 weeks of gestation and 36.5% between 27 and 36 weeks. A shift in medical care was observed at 26 weeks, with an increase in caesarean sections before labour and newborn referrals. In all, 63.3% of neonatal deaths were due to perinatal conditions, and 27.9% to congenital anomalies. The proportion of deaths explained by congenital anomalies was higher for longer gestational age: 14% of deaths between 25 and 28 weeks of gestation vs 38 to 43% between 33 and 42 weeks. CONCLUSION: The neonatal death certificate was well accepted; however the data on detailed causes of death and parent's characteristics were insufficient. Analysis of the circumstances and the causes of death is facilitated with the neonatal death certificate and it will be developped in the future.

[Ovarian cancer: incidence and mortality in France]. / Les cancers de l'ovaire: incidence et mortalité en France.

OBJECTIVES: To describe the ovarian cancer incidence and mortality trends in France from 1980 to 2000. MATERIALS AND METHODS: Incidence data were obtained from 9 French departments covered by cancer registries which systematically record all cancers, of which those of the ovary. Mortality data has been provided by INSERM bureau in charge of the analysis of death certificates. RESULTS: With 4500 incident cases and 3500 deaths for the year 2000 in France, ovarian cancer still shows a poor prognosis. Age adjusted incidence and mortality rates are almost stable over the study period at an annual rate of 9 by 100000 for incidence and 5.5 by 100000 for mortality. However the risk of developing this cancer and the risk of dying from it, steadily decreased for the cohort born in 1930 to 1950, and the beginning of this reduction corresponds exactly to the beginning of oral contraception. We present comparisons between departments and analyze the border-line tumors in parallel with invasive cancers. CONCLUSION: The cancer of the ovary represents only a small part of female cancers (3.8%) in France and its incidence is stable over time, but its prognosis is very bad.

Mortality from rheumatoid arthritis in France, 1970-1990.

BACKGROUND: The degree to which rheumatoid arthritis (RA) influences life expectancy and mortality remains controversial. There have been few attempts to analyse death certificate data for this condition. Despite limitations, the information derived from detailed death certificate analysis for a large population over a long period allows the examination of aspects of the disease process and its impact. METHODS: The mortality related to RA in France was investigated for 1970-1990 inclusive. A multiple cause of death analysis was conducted, based on information recorded in death certificates. RESULTS: Death from RA represented 0.22% of all deaths. Mortality due to RA was strongly influenced by sex (female/male ratio of number of deaths = 3.3) and age (proportional mortality ratios higher in the 65-74 and 75-84 year age groups). In women, the mean age at death for RA was slightly lower than the mean age at death from all causes, while the difference was clearly opposite in men. The impact of RA remained relatively constant during the study period, both in terms of proportional mortality and age at death. The analysis of associated causes of death did not yield significant changes in the pattern of death from RA. CONCLUSIONS: The study demonstrates the significant and relatively constant impact of RA on mortality in France over the study period. This suggests that the introduction of new treatment regimens or other environmental factors has had little influence on the impact of RA mortality in the community.

European study of the certification and coding of causes of death of six clinical case histories of diabetic patients. EURODIAB Subarea C Study Group.

This study was designed to investigate the large differences in diabetes mortality rates in Europe. In each of the participating countries (France, Germany, The Netherlands, Northern Ireland-UK, Republic of Ireland, Romania, Scotland-UK, Switzerland) a random sample of certifying physicians was asked to certify the causes of death of six case histories which described the deaths of diabetic patients; the responses from an average of 220 physicians per country were analysed. These registered causes were then coded nationally and the underlying cause was compared with that following a central recoding. Overall 28% of the physicians surveyed recorded diabetes on the death certificate as the underlying cause of death--France was 25% below this overall average and Germany 21% above. The national coding of diabetes as the underlying cause of death differed from the central recoding with a comparative undercoding of almost 40% in Romania, 30% in Northern Ireland and 25% in Switzerland; in contrast, there was an overcoding of diabetes by 80% in The Netherlands and 60% in the Republic of Ireland. After adjusting for central recoding, in part an adjustment for certification habits, the national coding from this simulation study was able to explain 35% of the variation in the diabetes mortality rates. With such differences in the coding of diabetes, the currently published mortality rates for diabetes are not directly comparable between European countries; some suggestions are made for the reduction of the intercountry differences in the collection and analysis of mortality data for diabetes.

Death certificate coding practices related to diabetes in European countries--the 'EURODIAB Subarea C' Study.

The objective of this study was to compare and analyse coding practices for diabetes mortality data in nine European countries (Belgium, Republic of Ireland, France, Germany, Malta, The Netherlands, Northern Ireland, Scotland and Switzerland). In each country, a sample of 200 coded death certificates, which mentioned diabetes, was randomly sampled. All death certificates were recoded at the WHO Collaborating Centre for the Classification of Diseases in the French language. The results show wide differences between national coding and central coding. Discrepancies in the underlying cause of death existed at the 3-digit coding level for 26% of all death certificates and for 44% at the 4-digit level. Coding in Northern Ireland and Malta was characterized by a marked tendency to choose diabetes less frequently. In contrast, in The Netherlands and, to a lesser extent, in the Republic of Ireland and France, diabetes was more frequently selected as the underlying cause of death. Most of the differences concerned the coding of an association involving diabetes and circulatory system diseases. In some countries, these coding differences influence the reported level of diabetes mortality. For Northern Ireland and Malta, the number of certificates with diabetes as the underlying cause of death was more than doubled after central recoding and for The Netherlands, in contrast, it was almost halved. To explain the differences a number of factors are considered: a lack of information from the International Classification of Diseases (ICD), on the application of the coding rules, between-country differences in cause of death certification practices, a divergence of opinion about the causal role of diabetes when it is associated with other conditions, a lack of homogeneity between countries in data collection procedures.(ABSTRACT TRUNCATED AT 250 WORDS)

Health and employment of a female population in an urban area.

The authors discuss the results of a survey of women aged 25 to 50 living in a university town in southern France (Montpellier). Morbidity of these women (reported health problems, consultation with health care professionals etc) was analysed as a function of whether or not they were employed out of the home. Overall, differences were minimal; after correction for demographic factors, the only significant difference was more frequent reports of fatigue, overwork or nervousness in the employed women. Satisfaction with life situations was more closely correlated with health than was employment per se. A great similarity was seen between, on the one hand, employed women satisfied with their occupational conditions and housewives who did not report boredom, and on the other hand, between employed women dissatisfied with their occupational conditions and housewives who reported boredom.

Sudden infant death and immunization: an extensive epidemiological approach to the problem in France--winter 1986.

In March 1986 five sudden infant deaths were reported, following the diphtheria-tetanus toxoids-pertussis and inactivated poliomyelitis virus (DTP-IPV) immunization of the infants concerned. An epidemiological study was carried out in order to investigate the possibility of a relationship between this immunization and sudden infant death syndrome (SIDS). A detailed examination of the five cases had been carried out by a doctor. An exhaustive survey of all postneonatal deaths occurring between January and March 1986 was conducted and also a matched case-control survey. No significant differences were found in the immunization rates between SIDS and other causes of death, nor between SIDS and living controls. These results are compared with the results from previously published studies on the topic.

AIDS-related conditions: study of a representative sample of 1203 patients deceased in 1992 in France.

BACKGROUND: Little representative information exists on the frequency of human immunodeficiency virus (HIV)-related diseases among the overall AIDS population. The objective of this research is to assess the nature, frequency and characteristics of these diseases among AIDS patients during their last year of life and to analyse these frequencies according to the mode of transmission and other socio-demographic and medical characteristics. METHODS: To obtain comprehensive data, we conducted an investigation based on retrospective collection of clinical information on a representative sample (1203 deaths) of all AIDS deaths that occurred in France during 1992. RESULTS: The frequency of the diseases was markedly higher than the one described in the AIDS surveillance registers and varied between homosexuals and intravenous drug users (IVDU). After controlling for other variables (age, CD4 counts, survival times) by means of logistic regression, homosexuality remained a significant explaining factor for Kaposi's sarcoma, cytomegalovirus infections, herpes simplex and cryptosporidiosis. In contrast, HIV encephalopathy, hepatitis, mental disorders, invasive candidiasis and cachexia were more frequent in male IVDU. Few differences were observed by sex. CONCLUSIONS: Several factors may explain the differences: variation in exposure to infectious agents, general health status, use of medical care and direct influence of the mode of HIV transmission. These data are of particular value for medical services in planning the magnitude of health care needs among the AIDS population overall, for clinicians and researchers for advancing the understanding of the natural history of AIDS and in the definition of prophylactic strategies against opportunistic infections.

"Avoidable" mortality and health services: a review of aggregate data studies.

STUDY OBJECTIVE: The aim of the study was to review published work reporting mortality from conditions amenable to medical intervention and compare the methods used and the results obtained. SOURCE MATERIAL: Two types of analysis were examined: (1) analyses of time trends, relating decline in mortality from amenable conditions to improvements in medical care (3 papers); (2) analyses of geographical variation, either between or within countries, in which mortality was related to the availability of health care resources and to other factors (8 papers). RESULTS: Time-trend studies have in general shown that mortality from amenable causes has declined faster over the past decades than most other causes of death. Studies of geographical variation have shown that mortality from amenable causes is consistently associated with socioeconomic factors, and that the association with the provision of health care resources is rather weak and inconsistent. CONCLUSIONS: (1) The low levels of mortality from amenable causes which presently prevail in industrialised countries are likely to reflect, at least in part, the increased effectiveness of health services; (2) geographical variation in mortality from amenable causes has not yet been shown to reflect differences in effectiveness of health services; and (3) if geographical variation in avoidable mortality does reflect such differences, they must arise from circumstances other than the level of supply, for example from more specific aspects of health care delivery, and are probably closely related to socioeconomic circumstances. In depth studies at the individual level are now more likely to produce information about factors limiting the effectiveness of health services than further studies of aggregate data.