Parkinson's disease (PD) is associated with a significant burden of illness and cost to society, which has been difficult to quantify. Our objective was to use linked administrative databases from the population of Ontario, Canada, to assess the prevalence of parkinsonism, physician- and drug-related costs, and hospital utilization for parkinsonian patients compared with age/sex matched controls. An inception cohort of parkinsonian cases from 1993/1994 was age and sex matched (1:2) to controls and followed for 6 years. Patients were identified by the diagnostic code for PD, the use of specific PD drugs, or a combination. The parkinsonian case cohort (15,304) was matched to (30,608) controls that did not have parkinsonism. The age-adjusted prevalence rates were 3.63 for men and for 3.24 women per 1,000 (increased by 5.4% for men and 9.8% for women). Physician costs were 1.4 times more, there were 1.44 times more hospital admissions, admissions were on average 1.19 times longer, and drug costs were 3.0 times more for parkinsonian cases. We conclude that the substantially higher physician and drug costs as well as hospitalization rates compared with controls clearly suggest that parkinsonism is associated with large direct costs to society.

Hypertension and tachycardia in the PACU, although infrequent, are associated with increased risk of unplanned critical care admission and mortality. Patient, surgical, intraoperative, or PACU observations contribute more to cardiovascular events in the PACU than do differences in anesthetic management identified in this study.

To study comorbidity in patients with Parkinsonism (PKM), relative hospitalization rates from 1994 to 1999 for 15,304 cases were compared with 30,608 controls. After correction for differential survival, the rates were higher for cases compared to controls for aspiration pneumonia (6.34; 95% confidence interval [CI], 5.23, 7.93), affective psychosis (2.71; 95% CI, 2.13, 3.32), hip fractures (2.56; 95% CI, 2.35, 2.76), other urinary tract disorders including infections (2.5; 95% CI, 2.17, 2.86), septicemia (2.39; 95% CI, 2.02, 2.85) and fluid and electrolyte disorders (2.27; 95% CI, 1.93,2.66). The rates for cardiac, cerebrovascular, and peripheral vascular disease were similar. Preventive measures and aggressive management of these conditions as outpatients may reduce the rates of hospitalization and improve the morbidity and mortality of PKM.

Most studies of postoperative nausea and vomiting have concentrated on single etiologic factors and have not detailed the method of assessing these symptoms. This study used postoperative interview data from patients at four teaching hospitals during 1988-89, to determine 1) risk factors for nausea/vomiting, 2) whether the type of surgery affected the rate of nausea/vomiting among female patients, 3) whether differences in rates across hospitals were due to differences in patient case-mix, and 4) whether there were differences in the rate of nausea/vomiting among the patients of individual anesthesiologists. Research nurses performed 16,000 interviews (59% of all inpatients) from a closed-question standardized format. With a multiple logistic regression that controlled simultaneously for all risk factors, factors associated with increased risk for nausea/vomiting for all patients included younger age, female, lower physical status score, no preoperative medical conditions, nonsmokers, elective procedures, longer duration of anesthesia, inhaled anesthetics, use of intraoperative opioids, and gynecologic or ophthalmologic operations. Among women, risk factors were similar, with minor gynecologic surgery associated with increased risk (relative odds = 2.30). We found marked variations in the rate of nausea/vomiting across hospitals (range, 39% to 73%), and these variations were not explained by the case-mix of patients. The rate of nausea/vomiting varied substantially across anesthesiologists in each hospital and the differences were not explained by differences in the patients they managed. Thus in the time period immediately preceding the introduction of newer antiemetic drugs, we found that the rates of this common problem were persistently high as perceived from the patients' point of view.

Despite modern drug therapy, PKM continues to confer a sharply increased mortality on unselected patients followed for several years.

Reporting health data for large urban areas presents numerous challenges. In the case of Toronto, Ontario, amalgamation in 1998 merged six census subdivisions into one megacity, resulting in the disappearance of standard reporting units. A population-based approach was used to define new health planning areas. Census tracts were used as building blocks and combined according to residential income homogeneity, respecting natural and man-made boundaries, forward sortation areas and the City of Toronto's community neighbourhoods whenever possible. Correlations and maps were used to establish area boundaries. The city was divided into 5 major planning areas which were further subdivided creating 15 minor areas. Both major and minor areas showed significant differences in population characteristics, health status and health service utilization. This commentary demonstrates the feasibility and describes the outcomes of one method for establishing planning and reporting areas in large urban centres. Next steps include the further generation of health data for these areas, comparisons with other Canadian urban areas, and application of these methods to recently announced Ontario Local Health Integration Networks. These areas can be used for planning and evaluating health service delivery, comparison with other Canadian urban areas and ongoing monitoring of and advocacy for equity in health.La traduction du résumé se trouve à la fin de l'article.