The Australian National Collaborative Research Infrastructure Strategy supports development of a national research capability in population health and clinical data linkage. This paper illustrates ...the importance of incorporating a population registry within such a system using an example provided by the Manitoba Injury Outcome Study (MIOS) that quantified the long‐term burden of mortality attributable to injury in working‐age adults.
MIOS is a population‐based matched cohort study that used administrative health data from Manitoba, Canada. An inception cohort of injured cases (ICD‐9‐CM 800–995) aged 18–64 years was identified from all Manitoba hospital admissions between 1988 and 1991. A matched non‐injured comparison group was randomly selected from the total provincial population using the Manitoba Population Registry. Mortality outcomes were obtained by linking the two cohorts with the deaths data over 10 years. Mortality rate ratios (MRRs) were calculated to compare the injured and non‐injured cohorts.
A total of 21,032 matched pairs were identified. Using the population registry, the 10‐year adjusted all‐cause MRR comparing injured and non‐injured cohort was 1.80 (95% CI 1.65–1.98). Without the registry, the unadjusted standardised morality ratio was 2.76 (95% CI 2.52–3.02).
The effect of injury on mortality outcomes was over‐estimated using only the injured cases, without use of the population registry. Use of the population registry enabled the selection of a matched non‐injured group for comparison purposes, ensured comprehensive follow‐up of almost all participants, and provided more accurate estimates of exposure time, incidence of mortality and relative risk.
Incidence and prevalence of diabetes in children aged 0-14 years in Manitoba, Canada, 1985-1993.
J F Blanchard ,
H Dean ,
K Anderson ,
A Wajda ,
S Ludwig and
N Depew
Manitoba Diabetes Education ...Programme, Manitoba Department of Health, Winnipeg, Canada.
Abstract
OBJECTIVE: To estimate the incidence and prevalence of type I diabetes among Manitoba children aged 0-14 years from 1985-1993.
RESEARCH DESIGN AND METHODS: The Manitoba Diabetes Database (a population-based database of individuals diagnosed with diabetes
based on Manitoba's health insurance system) was used to estimate the annual incidence of diabetes for the years 1985-1993
and the point prevalence of diabetes at 31 March 1993 for Manitoba children aged 0-14 years. The Diabetes Education Resource
for Children and Adolescents program database was used to correct incidence and prevalence rates for ascertainment using the
two-source capture-recapture method. RESULTS: The overall ascertainment rate of the Manitoba Diabetes Database was 95% for
incident cases and 93% for prevalent cases. The average annual incidence was 20.4 per 100,000 for children aged 0-14 years.
The annual incidence appears to be stable for all age-groups 0-14 years over the past decade. The point prevalence of diabetes
among children was 120.4 per 100,000. CONCLUSIONS: The incidence of type I diabetes in children aged 0-14 years in Manitoba
is higher than reported previously in other urban regions of Canada, but similar to population-based estimates from Prince
Edward Island. The incidence appears stable in Manitoba over the past decade even in the 0-4 year age-group. The Manitoba
Diabetes Database appears to be a highly accurate population-based source of data on the epidemiology of diabetes in children.
To determine, in a defined population, the percentage of persons who were discharged from a hospital or died of influenza-associated respiratory conditions who had a health care contact during the ...preceding vaccination season and to determine the relation between risk status for influenza-associated hospitalization and death and influenza vaccination rates.
An observational study using linked-record analysis of medical claims data.
A probability sample of 100,000 noninstitutionalized adults living in Manitoba in 1982 to 1983.
Analysis of medical claims for influenza vaccination and hospital discharges and deaths for influenza-associated respiratory conditions during the 1982-83 influenza vaccination season and influenza outbreak period.
For the population as a whole, 50% to 60% of elderly persons (greater than or equal to 65 years of age) and 30% to 40% of younger persons had one or more health care contacts during the influenza vaccination season but fewer than 10% of all persons had been discharged from a hospital. In contrast, for elderly persons hospitalized with respiratory conditions during the influenza outbreak period, approximately 80% had at least one health care contact during the vaccination season. Among the elderly, 39% to 46% of all those discharged for influenza-associated respiratory conditions and 62% to 67% of those who died had been discharged from hospital during the previous vaccination season. Persons discharged with high-risk conditions during the vaccination season were at greater risk for hospitalization with influenza-associated respiratory conditions but were less likely to be vaccinated than were those at lower risk.
Most persons who were hospitalized with influenza-associated respiratory conditions had contact with health care providers during the preceding influenza vaccination season. Among elderly patients, previous hospital care was common, especially among those who died. The disparity between influenza vaccination rates and risks for influenza-associated hospital discharge and death supports a strategy of hospital-based influenza vaccination.
Reported increases in the number of fractures of the proximal femur in Europe are greater than can be explained by demographic changes alone. This trend was assessed in Canada by examining hospital ...discharge records from the provinces of Saskatchewan and Manitoba from 1972 to 1984. The annual number of first fractures of the proximal femur in persons older than 50 years of age increased 59.7% in women and 42.2% in men during this time period. In most of the five-year age groups the percentage of increase in the number of fractures exceeded the percentage of increase in population of that age group. Annual age-specific incidences (by five-year age groups) increased exponentially with age, doubling every six years, and reached a maximum value of 4% in women older than 90 years of age. Annual age-adjusted incidences increased significantly over the study period in men and women. For the whole of Canada in 1987, it is estimated that there were 13,193 first fractures of the proximal femur in women and 4610 in men, and that in the year 2006 these will rise to 22,922 and 7846, respectively. The actual increase will be considerably greater if the age-specific incidences continue to increase as they have from 1972 to 1984. The gradual decline in physical activity, which contributes to bone loss, may be one etiological factor of this trend during the last half century.
With the growing reliance on large health care data bases, the need to verify data quality increases as well. Because of the considerable costs involved in checks using primary data collection, a ...computerized methodology for performing such checks is suggested. The technique seems appropriate for any situation where two data collection systems (i.e. hospital discharge abstracts and physician claims for payment) relate to the same event, such as a patient's hospitalization. After reviewing other approaches, this paper suggests linking physician claims for performing particular surgical procedures with hospital discharge abstracts for the stay in which the surgery took place. Physician and hospital data for adults age 25 and over in Manitoba from 1 April, 1979 to 31 March, 1984 were used to address the questions: 1. How well can the two data sets be linked? 2. Given linkage of the two data sets, how much agreement is there as to procedure and diagnosis? Linkage between hospital and physician data was excellent (over 95%) for 5 out of 11 surgical procedures (hysterectomy, prostatectomy, total hip replacement, coronary artery bypass surgery, and heart valve replacement); there was over 90% perfect agreement for three other procedures (cholecystectomy, cataract surgery and total knee replacement). Problems with matching the Manitoba Health Services Commission tariffs (on physician claims) with ICD-9-CM operation codes (on hospital data) led to only 77% perfect agreement for vascular surgery and 84% for gallbladder and biliary tract operations other than cholecystectomy; over 10% of the cases linked on surgeon and date but not on the designated procedures.
Claims-based indices of comorbidity and severity, as well as other measures derived from routinely collected administrative data, are developed and tested. The extent to which risk adjustments using ...claims can be improved by adding information from one well-known measure based on chart review and patient examination (the American Society of Anesthesiologists' (ASA) Physical Status score) is also examined. Readmissions and mortality after three common surgical procedures are the outcomes studied using multiple logistic regression. Claims-based measures of comorbidity, derived both from hospital discharge abstracts at the time of surgery and from hospitalizations in the 6 months before surgery, provided reasonably good predictions of postsurgical readmissions and mortality. In the most complete logistic regression models, the Somers' Dyx measure of fit (a rank correlation coefficient) ranged from 0.23 to 0.38 for readmissions and from 0.46 to 0.72 for mortality. In 5 out of 6 cases, these predictions were not improved by including the prospectively-collected ASA Physical Status score. Such difficulties in improving risk adjustment by more intensive data collection are discussed in terms of their research implications.
By understanding the range of approaches implicit in modern record linkage, epidemiologists and health services researchers can better decide its suitability for their needs. The authors discuss a ...small record linkage project, providing a sense of where mistakes were made. The research first uses existing identification numbers as a gold standard for linking hospital abstracts and physician claims to investigate whether or not coronary angiography was performed on a given individual. Even if identification numbers are not available, a successful linkage (with more than 95% of the cases matched) may be possible under some circumstances. The linkage process highlights problems with the consistent recording of coronary angiography in inpatient and outpatient hospital abstracts. Our approach should prove useful when the same procedure is recorded in more than one place on a single file and when validating a procedure (or other event) across files is important. Given the growing number of health care databases and ongoing changes in the delivery of care, record linkage often can provide quality control and expand research opportunities in a timely fashion.
PDF
