Study design & population
This was a retrospective cohort study designed to use Ontario’s provincial healthcare data to study the healthcare utilization of Veterans entering the provincial healthcare system in Ontario. A description of this cohort has been published [Mahar et al., Journal of Military, Veteran and Family Health, in press]. This study is a descriptive analysis of the non-mental health services utilization of the Canadian Veteran population in Ontario. Ethics approval for this study was obtained from the Queen’s University Health Sciences and Affiliated Teaching Hospitals Research Ethics Board. Individual patient consent was not required. The Institute for Clinical Evaluative Sciences is a s. 45 Prescribed Entity under Ontario’s privacy law (PHIPA) enabling us to study the health and health outcomes of individuals for the purpose of analysis or compiling statistical information with respect to the management of, evaluation or monitoring of, the allocation of resources to or planning for all or part of the health system.
Individuals were eligible for the study if they had registered for Ontario Health Insurance Plan (OHIP) coverage between January 1, 1990 and March 31, 2013 and provided evidence of a previous career in the CAF or RCMP. The date of OHIP registration was used as the index date for entry into the cohort and approximated the service release date. Individuals were followed from the time they left the CAF or RCMP until death, the end of OHIP eligibility (e.g., moved out of province), or the end of the follow-up period (March 31, 2013).
Identifying veteran status
In Canada, the healthcare of Canadian Armed Forces (CAF) personnel and RCMP officers is federally regulated. CAF personnel are provided care within a specialized military system separate from the provincial healthcare systems. Until April 1, 2013, RCMP officers were considered non-provincial residents and their health services use was billed to the federal government [1–3]. In Ontario, the Ministry of Health and Long Term Care (MOHLTC) tracks registration for provincial healthcare coverage for previous CAF and RCMP members using specific administrative codes on the OHIP application form and this is stored alongside the health card number. This form provides the necessary documentation on service history to waive the standard three-month waiting period for provincial healthcare coverage. For this study, Veterans were defined as CAF and RCMP service leavers who provided evidence to the MOHLTC about their career history. The MOHLTC provided the authors with the anonymized list of people with an administrative CAF and RCMP service code linked to their health card number, as well as career start and end dates. This information was linked to the Institute for Clinical Evaluative Sciences (ICES) data holdings using unique encoded identifiers and analyzed at ICES. Details on cohort creation and demographics are provided elsewhere [8, 9].
Data sources & variables
The following administrative healthcare databases housed at ICES were linked to the MOHLTC Veteran data to describe the provincial health services utilization of Ontario Veterans over time. The Canadian Institute for Health nformation Discharge Abstract Database (CIHI-DAD), the National Ambulatory Care Reporting System (NACRS), OHIP and the Registered Persons Database (RPDB).
The CIHI-DAD is a repository of inpatient admission and discharge data (except hospitalizations occurring in designated psychiatric beds as of 2005), including all medical inpatient admissions in the province. It was used to identify non-mental health related hospitalizations and cumulative inpatient stay. The NACRS database is a warehouse of outpatient hospital visit data, including emergency department, oncology and dialysis. It was used to identify non-mental health related emergency department (ED) visits. The OHIP database is a repository of physician billing data for healthcare encounters and procedures. It was used to identify non-mental health related family physician visits, as well as to determine provincial healthcare coverage eligibility dates for the cohort. Standard ICES definitions were applied to the data to measure these healthcare encounters and preventative health behaviors . Physician and emergency department visits, and hospitalizations related to mental disorders will be reported elsewhere. The Registered Persons Database (RPDB) provided demographic information, using data supplied by the MOHLTC and supplemented by other provincial data holdings at ICES.
Non-mental health services utilization
In this study, we exclusively evaluated medical and non-mental health services utilization among Veterans in Ontario. Non-mental health hospitalizations were defined as admissions that did not have a most responsible diagnosis of a mental disorder (e.g., all ICD-10 F codes). They were measured as a dichotomous variable (hospitalized yes/no) and as a count (number of hospitalizations). Cumulative inpatient stay was defined as the total number of days spent in hospital, during the study timeframe, regardless of the number of re-admissions or location of stay. ED visits were measured as a dichotomous variable (ED visit yes/no) and as a count (number of ED visits). To measure non-mental health related ED visits, visits with a most responsible diagnosis of a mental disorder (all ICD-10 F codes) were excluded. Family physician and specialist physician visits were studied separately. Family physician visits were restricted to defined as visits to doctors with specialties in family medicine, family medicine/emergency medicine. Specialist visits were defined as visits with any other specialist, excluding psychiatry. Both were measured as dichotomous variable (yes/no) and count (number of visits). Mental health related family physician visits were excluded according to a previously validated algorithm [10, 11]. Healthcare encounters with physicians for mental disorders were also excluded from this description using ICD-9 codes.
A description of all resource utilization measures is presented using frequencies and 95 % confidence intervals for categorical variables. One-sample proportion confidence intervals for categorical variables were calculated using the normal approximation to the binomial distribution. Mean, standard deviation (SD), median and interquartile range (IQR) are provided for continuous and count data.
All descriptive statistics were stratified by five year time intervals following release (0–5 years, 5–10 years, 10–15 years, and 15–20 years) to investigate temporal patterns in medical healthcare utilization within the cohort. Baseline was considered the first five years following release. This is similar to how the results of a panel study are presented, integrating both cross-sectional and cohort designs to present snapshots of a cohort over time . To determine if differences existed across time, 95 % confidence intervals were used, corresponding with a p <0.05.
We a priori anticipated that a cohort effect by age at release would exist. To explore possible patterns in medical healthcare utilization, stratified descriptive statistics were also presented by age at release categories. To determine if differences existed across age categories, 95 % confidence intervals were used, corresponding with a p <0.05. Cell sizes less than 6 are suppressed according to ICES privacy regulations. All descriptive analyses were performed using SAS 9.3 (Cary, North Carolina, USA).