Return to search

Magnitude and causes of small area variation in surgical rates of myringotomy with the insertion of ventilation tube in Ontario

Otitis media is often treated surgically through the use of a procedure termed myringotomy with insertion of a ventilation tube (MVT). From fiscal year 1992/3 to 1997/8, some 132,000 such operations were performed in Ontario, Canada, on young people < 20 years old. This study quantifies the small area variation (SAV) in MVT surgical rates across the 49 counties and 478 postal areas of Ontario; examines the rate variation as a function of variations in physician practice style, physician supply, physician demographic characteristics, urbanrural differences, and population socio-economic status; and identifies economic and policy-related implications of SA V. Physician practice style was assessed using responses from a comprehensive provincial survey of referring physicians (OPs and paediatricians) and specialists/surgeons concerning their opinions of the indications for, and treatment of, otitis media, with particular reference to MVT surgery. Other data were obtained from physician databases and Canada census statistics. The variation in age-gender standardized MVT rate across the counties is highly statistically significant, as demonstrated by Chi-square tests, and by comparison of simulated and observed values of variation statistics (the EQ, CV, and SCV) that show much higher inter-area variation than that expected by chance alone. Multivariate regression modelling of inter-county MVT rate variation across the counties showed six variables to be significant determinants of surgical rate. MVT rate is a positive function of physician propensity to refer patients to surgeons, the percentage of North American-trained referring physicians, and the supply of paediatricians; and a negative function of the supply of surgeons, and the median licence years of referring physicians and surgeons respectively. These variables explain 74 % of the SA V in MVT rate for the 35 counties for which the regression model applies. Over 40 % of the explained variation is accounted for by differences across the counties in the propensity of physicians to refer patients to surgeons. Socio-economic status (as measured by an index of deprivation) was not a significant predictor of inter-county MVT rate. When examined at the FSA (postal forward sortation area) level within counties, the effect of socio-economic status was small overall but locally significant for five counties. Urban counties, although better equipped with medical care resources, including the supply of physicians, generally maintain lower MVT rates than rural counties. This is interpreted as being due to inter-county referral flow from rural GPs and paediatricians to specialists operating in urban counties, particularly those counties containing teaching centres and training hospitals. In addition, the higher surgical rates in rural areas are also associated with a higher percentage of North American-trained referring physicians working in such areas. The large regional variations in MVT rates, and the determinants of such variations, have implications for health care providers and policymakers. Between 1992/3 and 1997/8, the welfare loss resulting from MVT surgical rate variation that was due to variation in physician variables (physician opinion, demographics, and supply) amounted to around $7.5 Million per year (or about three-quarters of the annual MVT surgical spend). More than 40 % of the welfare loss was attributable to inter-county variation in physician propensity to refer patients to surgeons. This welfare loss could in theory be reduced by the production and dissemination of information such as clinical guidelines, aimed at modifying physician behaviour and reducing variation in physician practice style (in particular, reducing variation in physician propensity to refer). However, published findings indicate that such change strategies are generally ineffective in modifying physician behaviour.

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:250901
Date January 2000
CreatorsAsche, Carl V.
PublisherUniversity of Surrey
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttp://epubs.surrey.ac.uk/755037/

Page generated in 0.002 seconds