Can the Quality of Care in Family Practice Be Measured Using Administrative Data

摘要: Primary care is the foundation of Canadian health system. A strong Health Care (PHC) system has been shown to result in a healthier population (Shi 1997; Macinko, Starfield, and Shi 2003) may also contribute more than do specialized services (Starfield 2002; Baicker Chandra 2004). Indeed, most problems are initially treated by primary physicians (i.e., family or general practitioners; Green et al. 2001). An effective PHC necessary address challenges an aging population, increase chronic disease, complex comorbidity, and/or functional disability (Future Family Medicine Project Leadership Committee Recognition this need led efforts improve delivery several countries access quality. The Institute defined quality as “the degree which for individuals populations likelihood desired outcomes consistent with current professional knowledge” (Lohr 1990, p. 128–129). Reviews practice United States, Kingdom, Australia, New Zealand have found established standards be rarely met (Seddon 2001; McGlynn 2003). In Canada, we know comparatively little about care. effort this, Services Policy Research (CIHR) recently identified “Managing safety” priority area research funding. The expressed interest “research designed identify management strategies at affordable cost support extensive use performance indicators …” (Canadian Institutes 2004). Fundamental process development tools methods measure (Donabedian 1980). For example, there recent using Quality can measured terms structures (characteristics such personnel, equipment, finances), processes (the actual given encompassing clinical interpersonal effectiveness (Campbell 1998), consequences status user satisfaction; Donabedian While these three dimensions somewhat interdependent, good one does not imply another (Gandhi 2002). In care, appropriate measures because they depend on factors unrelated system, socioeconomic (Frohlich Mustard 1996; Sheldon 1998; Martens 2002), well upon provided all levels care—primary, secondary, tertiary. Instead, generally accepted useful (Brook, McGlynn, Cleary 1996); thus, study described article sought develop measures, focusing specifically physicians. These examined different approaches. Surveys allow collection data both satisfaction but costly subject recall nonresponse biases (Vogt Medical record audits provide comprehensive view expensive, limited poor documentation patient records (Marshall 2003; Vogt Using chart information combination available from administrative databases poses significant confidentiality environment where electronic charts uncommon. Direct observation used tool very expensive potentially intrusive (Stange 1998). Thus, limitations make unrealistic source longitudinal monitor Shekelle 2000). Administrative alternative cost-effective general, measurement indicators, particular. Although daily work running rather being collected purposes, readily each full year individual level, thereby allowing researchers examine important patterns “health service use, expenditures, selected outcomes, care” across various settings over time (Iezzoni, Shwartz, Ash 2005, 141). public private insurers only select groups people, according scope particular insurer. They vary content format. Canada's however, provides residents first dollar coverage medically physician hospital services. Each province responsible administration its within parameters set out Federal Canada Act. provincial stem single cover entire population. Manitoba, almost every contact recorded Province's database billing purposes (Black, Roos, Roos 2005). claims, retail pharmacies prescriptions dispensed community, abstracts submitted day surgery inpatient stays, Home program regarding receipt services, personal (nursing) home admissions. Physician claims fee-for-service (who submit Manitoba government remuneration), paid completely part via alternate payment mechanisms (e.g., salaried, contract); latter group (called shadow billings) purposes. Fee-for-service reliably provide, plans. Most (80 percent) bill (Katz 2004; Watson Optometrists entitled “medically necessary” eye exams diabetic patients. All other billed directly patient; appear data. Manitoba assigns unique numeric identifier person registered insurance that allows them tracked sectors longitudinally. Manitobans out-of-province separate file. However, “reciprocal” represent less 1 percent our focus was regular in-province residents, chose exclude claims. Hence, rich consistently long period lend themselves population-based research. Several provinces Canada—notably, British Columbia, Ontario, Quebec—are already engaged (Tamblyn 1995; Menec, Currie 2004). The larger study. goal explore feasibility measuring Manitoba. Our two key objectives were to: (a) acceptable practicing physicians; (b) describe indicators. objective.