Linkage Of Administrative Data Research Articles

Advantages and limitations of using national administrative data on obstetric blood transfusions to estimate the frequency of obstetric hemorrhages

Obstetric hemorrhages are a frequent cause of maternal death all over the world, but are not routinely monitored. Health systems administrative databases could be used for this purpose, but data quality needs to be assessed. Using blood transfusion data recorded in administrative databases to estimate the frequency of obstetric hemorrhages. Research design A population-based study. Subjects Validation sub-sample: all mothers who gave birth in a French region in 2006-07 (35 123 pregnancies). Main study: all mothers who gave birth in France in 2006-07 (1 629 537 pregnancies). Linkage and comparison of administrative data on blood transfusions with data from the French blood agency ('gold standard'), and, based on this validation, the construction of a multivariable regression model to correct the number of pregnant women identified as having received a transfusion in the national administrative database. The blood transfusion rate observed in the gold standard was 7.12‰. The sensitivity of the administrative data was estimated at 66.3% and the positive predictive value at 91.3%. The estimated total number of pregnant women who received blood transfusions in France in 2006-07 was 10 941 (6.71‰). The administrative data, available in most countries, can be used to estimate the frequency of obstetric hemorrhages.

Register-based studies of cardiovascular disease

The use of the unique personal identification number in the Nordic database systems enables the researchers to link the registers at the individual level. The registers can be used for both defining specific patient populations and to identify later events during follow-up. This review gives three examples within cardiovascular epidemiology to illustrate the use of the national administrative registers available to all researchers upon request. The hospitalisation rate of acute myocardial infarction (AMI) was expected to be increased and case-fatality rate to decrease when the diagnostic criteria were changed in 2000. Linkage of national registers found a relative increase in hospitalisation rate of 14% while the case-fatality rate was unaffected. The pharmacological treatment of AMI patients was evaluated by linkage of administrative data. The use of evidence-based treatment increased significantly over time and adherence to treatment was high. Finally, use of specific nonsteroidal antiinflammatory drugs by healthy subjects was associated with a dose-dependent increase in cardiovascular risk. The nationwide registers have proven very useful in monitoring the hospitalisation rate and treatment of cardiovascular disease. The risk of unmeasured factors affecting the results calls for cautious interpretation of the results.

Resource Use and costs of End-Of-Life/Palliative Care: Ontario Adult Cancer Patients Dying during 2002 and 2003

The objective of this study is to estimate the direct medical cost of end-of-life and palliative (EOL/PAL) care for cancer patients during the last six months of their lives--or, during the period from diagnosis to death, if briefer--in 2002 and 2003, in Ontario, Canada. A linkage of cancer registry and administrative data is used to determine the costs of health care resources used during the EOL/PAL care period. Costs are analyzed by cancer diagnosis, location of death, and type of service. The total Ontario Ministry of Health-funded cost of EOL/PAL care for cancer patients is estimated to be about CAD$544 million per year, with an average per patient cost of about $25,000 in 2002-2003. Our results suggest that acute care consumes 75 percent of EOL/PAL funding and that only a small proportion of health care services used by EOL/PAL care cancer patients is likely to be formal palliative care.

Understanding the determinants of consent for linkage of administrative health data with a longitudinal survey

The Survey of Families, Income and Employment (SoFIE) is a panel income study with a health component, including data linkage to national health databases. This paper investigates factors that predict consent to data linkage. SoFIE participants were asked permission to link their SoFIE records to administrative health data in Wave 3 of data collection. Logistic regression was used to investigate the association of sociodemographic variables with consent to record linkage. Of Wave 3 participants 79.4% consented to health record linkage. Taking into account prior attrition, consent and matching the overall participation rate was 60.9%. Participants reporting Māori, Pacific and Asian ethnicity had lower odds of consenting. There was a strong relationship between the presence of co-morbid disease and consent to linkage. This study has shown that consent to link health information to survey records was high and the survey provides a rich and unique dataset for longitudinal analyses.

Linking a pharmaceutical claims database with a birth defects registry to investigate birth defect rates of suspected teratogens

Data linkage of population administrative data is being investigated as a tool for pharmacovigilance in pregnancy in Australia. Records of prescriptions of known or suspected teratogens dispensed to pregnant women have been linked to a birth defects registry to determine if defects associated with medicine exposure can be detected. The Pharmaceutical Benefits Scheme is a national claims database that has been linked with population-based data to extract linkages for women with a pregnancy event in Western Australia from 2002 to 2005 (n = 106 074). Records of births to the women who were dispensed medicines in categories D or X of the Australian ADEC pregnancy risk category were linked to the Birth Defects Registry of Western Australia. Population rates of registered birth defects per 1000 births were calculated for each medicine. There were 47 medicines dispensed at least once during pregnancy with 23 associated with a registered birth defect to a woman dispensed the medicine. When the birth defect rate for each medicine was compared with the rate for all other women not dispensed that medicine, most medicines showed an increased risk. Medicines with the higher risks were medroxyprogesterone acetate (OR: 1.8; 95%CI: 1.4-2.3), follitropin alfa (OR: 2.5; 95%CI: 1.2-5.0), carbamazepine (OR: 3.1; 95%CI: 1.7-5.6) and enalapril maleate (OR: 8.1; 95%CI: 1.6-41.7). Many known associations between medicines and birth defects were identified, suggesting that linked administrative data could be an important means of pharmacovigilance in pregnancy in Australia.

Hospitalized prevalence and 5-year mortality for IBD: Record linkage study

To establish the hospitalized prevalence of severe Crohn's disease (CD) and ulcerative colitis (UC) in Wales from 1999 to 2007; and to investigate long-term mortality after hospitalization and associations with social deprivation and other socio-demographic factors. Record linkage of administrative inpatient and mortality data for 1467 and 1482 people hospitalised as emergencies for > or = 3 d for CD and UC, respectively. The main outcome measures were hospitalized prevalence, mortality rates and standardized mortality ratios for up to 5 years follow-up after hospitalization. Hospitalized prevalence was 50.1 per 100 000 population for CD and 50.6 for UC. The hospitalized prevalence of CD was significantly higher (P < 0.05) in females (57.4) than in males (42.2), and was highest in people aged 16-29 years, but the prevalence of UC was similar in males (51.0) and females (50.1), and increased continuously with age. The hospitalized prevalence of CD was slightly higher in the most deprived areas, but there was no association between social deprivation and hospitalized prevalence of UC. Mortality was 6.8% and 14.6% after 1 and 5 years follow-up for CD, and 9.2% and 20.8% after 1 and 5 years for UC. For both CD and UC, there was little discernible association between mortality and social deprivation, distance from hospital, urban/rural residence and geography. CD and UC have distinct demographic profiles. The higher prevalence of hospitalized CD in more deprived areas may reflect higher prevalence and higher hospital dependency.

Use of Administrative Data for the Surveillance of Mood and Anxiety Disorders

There is increasing interest in the use of administrative data for surveillance and research in Australia. The purpose of the present study was to evaluate the usefulness of such data for the surveillance of mood and anxiety disorder using databases from the following Canadian provinces: British Columbia, Ontario, Quebec and Nova Scotia. A population-based record-linkage analysis was done using data from physician billings and hospital discharge abstracts, and community-based clinics using a case definition of ICD-9 diagnoses of 296.0-296.9, 311.0, and 300.0-300.9. The prevalence of treated mood and/or anxiety disorder was similar in Nova Scotia, British Columbia, and Ontario at approximately 10%. The prevalence for Quebec was slightly lower at 8%. Findings from the provinces showed consistency across age and sex despite variations in data coding. Women tended to show a higher prevalence overall of mood and anxiety disorder than men. There was considerably more variation, however, when treated anxiety (300.0-300.9) and mood disorders (296.0-296.9, 311.0) were considered separately. Prevalence increased steadily to middle age, declining in the 50s and 60s, and then increased after 70 years of age. Administrative data can provide a useful, reliable and economical source of information for the surveillance of treated mood and/or anxiety disorder. Due to the lack of specificity, however, in the diagnoses and data capture, it may be difficult to conduct surveillance of mood and anxiety disorders as separate entities. These findings may have implications for the surveillance of mood and anxiety disorders in Australia with the development of a national network for the extraction, linkage and analysis of administrative data.

Influence of Socioeconomic Status on Lifestyle Behavior Modifications Among Survivors of Acute Myocardial Infarction

The impact of secondary prevention initiatives on survival in higher-risk socioeconomically disadvantaged patients after acute myocardial infarction (AMI) may depend on behavioral adaptive responsiveness, uptake, and adherence to healthier lifestyles. From December 1999 to February 2003, 1,801 patients in Ontario, Canada were interviewed regarding their lifestyle behaviors at 30 days after their index AMI hospitalization. Data were obtained using self-reported surveys, medical chart abstraction, and administrative data linkage. Multivariate analyses were adjusted for baseline sociodemographic, cardiac risk severity, and co-morbid conditions. Socioeconomically disadvantaged patients had greater cardiac risk severity at baseline than did their wealthier better-educated counterparts. Compared with lower-income patients, patients with higher incomes were less likely to smoke (adjusted odds ratio [OR] for highest vs lowest income tertiles 0.36, 95% confidence interval [CI] 0.21 to 0.63, p <0.001), more likely to participate in exercise (adjusted OR 1.40, 95% CI 1.07 to 1.85, p = 0.02), and more likely to decrease or discontinue alcohol use (adjusted OR 1.64, 95% CI 1.16 to 2.34, p = 0.06). The relation between education and lifestyle behaviors was less pronounced for education than for income. After adjustment for baseline factors, patients who acknowledged participation in regular physical exercise at 1 month had a significantly lower long-term mortality than those who did not. In conclusion, socioeconomically disadvantaged patients were sicker at baseline and less behaviorally responsive to embarking on healthy lifestyle changes after AMI than were those of higher socioeconomic status.

Health-related behaviours in adults with diabetes: associations with health care utilization and costs.

The primary objective of this study was to examine whether physical activity, diet, and smoking behaviours are associated with health resource utilization and costs in the Canadian context. A secondary objective was to evaluate demographic and health behavioural characteristics of the participants of the study to assess the degree of respondent bias. Self-reported physical activity, diet and smoking status were obtained from a large population-based sample of adults with diabetes (N = 2311). Resource utilization and cost information was obtained by linking these data to the provincial government's administrative database. Multiple regression models examined predictors of resource utilization and costs for individuals with type 1 (T1D) and type 2 (T2D) diabetes separately. To assess the degree of responder bias, characteristics of individuals who consented to link data were compared with those who did not consent. Various measures of health care utilization and costs were negatively associated with physical activity behaviour in both T1D and T2D groups. Ever having smoked cigarettes was associated with higher resource utilization in individuals with T2D when controlling for demographic and health variables. Significant differences in demographic and health behavioural characteristics of the participants who provided consent for data linkage and those who did not were also found. These findings are of interest considering that PA is a critical but understudied component of individuals with diabetes, and this appears to be one of the first studies to directly examine the relationship between health-related behaviours and health care utilization and costs. The findings may be useful in guiding targeted health promotion programs for individuals with diabetes. The results also indicate that studies involving linkage of administrative and survey data could be over-represented by healthy individuals.

Health-related Behaviours in Adults with Diabetes

The primary objective of this study was to examine whether physical activity, diet, and smoking behaviours are associated with health resource utilization and costs in the Canadian context. A secondary objective was to evaluate demographic and health behavioural characteristics of the participants of the study to assess the degree of respondent bias. Self-reported physical activity, diet and smoking status were obtained from a large population-based sample of adults with diabetes (N=2311). Resource utilization and cost information was obtained by linking these data to the provincial government’s administrative database. Multiple regression models examined predictors of resource utilization and costs for individuals with type 1 (T1D) and type 2 (T2D) diabetes separately. To assess the degree of responder bias, characteristics of individuals who consented to link data were compared with those who did not consent. Various measures of health care utilization and costs were negatively associated with physical activity behaviour in both T1D and T2D groups. Ever having smoked cigarettes was associated with higher resource utilization in individuals with T2D when controlling for demographic and health variables. Significant differences in demographic and health behavioural characteristics of the participants who provided consent for data linkage and those who did not were also found. These findings are of interest considering that PA is a critical but understudied component of individuals with diabetes, and this appears to be one of the first studies to directly examine the relationship between health-related behaviours and health care utilization and costs. The findings may be useful in guiding targeted health promotion programs for individuals with diabetes. The results also indicate that studies involving linkage of administrative and survey data could be over-represented by healthy individuals.

The impact of different rehabilitation strategies after major events in the elderly: the case of stroke and hip fracture in the Tuscany region

BackgroundOn a regional level, our aims were to describe rehabilitation patterns for elderly patients with stroke and hip fracture and to investigate mortality risk during the 6-month post acute period.MethodsData sources included administrative data relative to patients aged 65+ resident in Tuscany admitted in hospital for stroke or hip fracture between 2001 and 2003, traced up to 3 years before and 6 months following index admission. The study design involves computerized linkage of administrative data, and an exploratory analysis of the association between rehabilitation patterns and 6-month mortality, adjusting for clinical, demographic, and acute-related care characteristics using multivariate Cox regression.ResultsRehabilitation patterns vary greatly across Tuscany with considerable cost implications. Six month mortality risk for stroke patients is significantly lower among residents of Local Health Authorities where patients are more frequently rehabilitated, specifically in extra-hospital settings.ConclusionOur study, targeting two crucial conditions for elderly patients, found a high variability of rehabilitation patterns across a region, albeit coherent between the two pathologies, associated with remarkable differences in average expenditure. Differences in hazard rates for 6-month mortality after stroke at population level were also found. These results need to be confirmed and further investigated through a more robust information framework.

Data Linkage Enables Evaluation of Long-Term Survival after Intensive Care

Outcomes of intensive care are important to the patient and for assessment of benefit. Short-term outcomes after critical illness are well described, but less is known about long-term outcomes. This study describes the use of data linkage, combining intensive care unit (ICU) clinical data with administrative morbidity and mortality data, to assess long-term outcomes after treatment in ICU. The hospital-based cohort study was conducted in a 22-bed general ICU in a metropolitan teaching hospital. All patient admissions admitted to ICU from 1 January 1987 to 31 December 2002 were included. The prospective ICU clinical database with patient demographics, ICU diagnoses, severity of illness, daily assessment of organ failures and common daily treatments used was linked using probabilistic methods to the state-wide hospital morbidity and mortality databases to describe long-term survival. There were 26,019 ICU admissions (22,980 patients) with 25,972 records (99.8%) linked to a hospitalization event that included the index ICU admission. Unadjusted survival was 84.7% at 1 year decreasing progressively to 50.7% at 15 years. Age, type of admission, severity of illness (measured by Acute Physiologic and Chronic Health Evaluation (APACHE) II and the presence of organ failure), ICU length of stay, comorbidity (Chronic Health Evaluation and Charlson comorbidity index) and ICU admission diagnosis, were all associated with survival at 1, 3, 5, 10, and 15 year follow-up (P<0.001 at all time points). Linkage of clinical and administrative data provides a feasible method for ascertaining long-term survival after critical illness. Age, admission severity of illness, diagnosis and comorbidity influenced long-term unadjusted survival.

Chronic kidney disease as a risk factor for cardiovascular disease and all-cause mortality: a pooled analysis of community-based studies.

Chronic kidney disease (CKD) is a major public health problem. Conflicting evidence exists among community-based studies as to whether CKD is an independent risk factor for adverse cardiovascular outcomes. After subjects with a baseline history of cardiovascular disease were excluded, data from four publicly available, community-based longitudinal studies were pooled: Atherosclerosis Risk in Communities Study, Cardiovascular Health Study, Framingham Heart Study, and Framingham Offspring Study. Serum creatinine levels were indirectly calibrated across studies. CKD was defined by a GFR between 15 and 60 ml/min per 1.73 m(2). A composite of myocardial infarction, fatal coronary heart disease, stroke, and death was the primary study outcome. Cox proportional hazards models were used to adjust for study, demographic variables, educational status, and other cardiovascular risk factors. The total population included 22,634 subjects; 18.4% of the population was black, and 7.4% had CKD. There were 3262 events. In adjusted analyses, CKD was an independent risk factor for the composite study outcome (hazard ratio [HR], 1.19; 95% confidence interval [CI], 1.07-1.32), and there was a significant interaction between kidney function and race. Black individuals with CKD had an adjusted HR of 1.76 (95% CI, 1.35-2.31), whereas whites had an adjusted HR of 1.13 (95% CI, 1.02-1.26). CKD is a risk factor for the composite outcome of all-cause mortality and cardiovascular disease in the general population and a more pronounced risk factor in blacks than in whites. It is hypothesized that this effect may be due to more frequent or more severe subclinical vascular disease secondary to hypertension or diabetes in black individuals.

WW2: USING LINKABLE CLINICAL AND ADMINISTRATIVE CLAIMS DATA TO IMPROVE OUTCOMES RESEARCH: SELECTED CASE STUDIES FROM ONCOLOGY AND DIABETES MELLITUS

OBJECTIVES: The objectives of this workshop are to describe various types of linkable clinical and administrative claims data, review selected case studies, and highlight the strengths and limitations of these databases. PARTICIPANTS WHO WOULD BENEFIT: Those involved in planning, designing, implementing, and using data from retrospective database studies would benefit from this workshop. Administrative claims are increasingly used for outcomes research studies, despite known limitations associated with data reliability and validity, and a lack of clinical content. In recent years, efforts have been made to link claims to other sources, such as disease registries and clinical laboratory files, to create richer databases for research purposes. These linked data sources offer the potential for improved accuracy in case identification and outcomes ascertainment. For example, study patients can be selected based on their presence in a disease registry instead of relying on diagnoses reported on medical claim forms, and clinical laboratory files can be used to evaluate the success or failure of therapy. In this workshop, we will review, via case studies from oncology and diabetes, the content of linkable clinical and claims databases, the specific ways in which such data have been used in published outcomes studies, and the remaining limitations of this record linkage approach. Participants will learn how clinical data can be applied to strengthen studies of treatment costs and the burden of illness, and they will gain an appreciation of the improvements that linked data sources can make to the pharmacoeconomics and outcomes research fields. One case study will review applications of the SEER-Medicare database, which includes a linkage between cancer registry data and Medicare administrative claims for approximately 14% of U.S. cancer cases across 17 diverse regions. The other illustration will describe published studies of the economic benefits of improving glycemic control among diabetes patients, in which claims data from several managed-care organizations were linked to glycosylated hemoglobin test results.