|
Resource use and survival for patients hospitalized with congestive heart failure: differences in care by specialty of the attending physician |
Auerbach A D, Hamel M B, Davis R B, Connors A F, Regueiro C, Desbiens N, Goldman L, Califf R M, Dawson N V, Wenger N, Vidaillet H, Phillips R S |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology Management, by cardiologists or generalists, of patients with congestive heart failure.
Economic study type Cost-effectiveness analysis.
Study population Patients (aged over 18 years) hospitalised with an exacerbation of congestive heart failure.
Setting Hospital. The economic study was conducted in Massachusetts, California, Wisconsin, North Carolina and Ohio, USA.
Dates to which data relate Effectiveness and resource use data were collected between June 1989 to June 1991, and between January 1992 to January 1994. 1994 prices were used.
Source of effectiveness data Effectiveness data were derived from a single multi-centre study.
Link between effectiveness and cost data Costing was undertaken prospectively on the same patient sample as that considered in the effectiveness analysis.
Study sample Power calculations were not reported as having been used to determine the sample size. 1,298 patients hospitalised with a primary diagnosis of acute exacerbation of congestive heart failure, and whose attending physicians were cardiologists or general internists, were considered. 743 patients with a mean (SD) age of 63.3 (14.1) years were patients of cardiologists with a mean age of 45.7 years, while 555 patients with a mean (SD) age of 71.4 (21.3) years were patients of generalists with a mean age of 40.3 years.
Study design A prospective, multi-centre (observational) cohort study (SUPPORT - Study to understand Prognoses and Preferences for Outcomes and Risks of Treatments) carried out in 5 geographically diverse institutions. Maximum duration of follow-up was 4.6 years. Loss to follow-up was not reported. Chart abstractors gathered information on the outcome measures.
Analysis of effectiveness The analysis was based on intention to treat. The main health outcomes considered in the analysis were:
(1) average daily Therapeutic Intervention Scoring System (TISS) score, a chart-derived additive measure of resource intensity which assigns one point for minor and two to four points for major interventions, for days 1-25 of hospitalisation;
(2) right-heart catheterisation during the first 7 study days;
(3) coronary angiography during hospitalisation;
(4) whether an echocardiogram or radionuclide (multiple-gated image acquisition) scan was obtained during hospitalisation;
(5) continuous electrocardiographic monitoring during the first 7 study days;
(6) transfer to an ICU from a floor setting;
(7) use of specific medication on discharge;
(8) survival censored at 30, 180 and 365 days; and
(9) survival at 31 December 1994.
Medical records and telephone follow-ups were used to determine survival to 6 months after enrolment. Compared with patients of generalists, patients of cardiologists were younger (mean age 63.3 versus 71.4, p<0.001) and had lower Acute Physiology Scores at the time of admission (35.1 versus 36.7, p<0.001). They were, however, more likely to have a history of ventricular arrhythmias (21.0% and 10.2%, p<0.001). Adjustment for socio-demographic characteristics and severity of illness was performed using regression analyses. Cox proportional hazards modelling was used to estimate relative hazards for death at all censoring times considered and to estimate adjusted survival. The validity of the study findings was investigated by using involved multiple subset and stratified analyses. A logistic regression model was used to generate a propensity score as an adjustment for the important confounding effect of difference in access to specialty care.
Effectiveness results The average daily TISS scores were 2.83 points higher for patients of cardiologists than for patients of generalists (CI: 1.96 - 3.68 points). Patients of cardiologists were more likely to undergo right-heart catheterisation (adjusted odds ratio 2.9 (CI: 1.7 - 4.9)) or cardiac catheterisation (adjusted odds ratio 3.9, (CI: 2.4 - 6.2)) and had higher odds for transfer to an intensive care unit and electrocardiographic monitoring. Adjusted survival did not differ significantly between groups at 30 days. However, there was a trend toward improved survival among patients of cardiologists at 1 year (adjusted relative hazard 0.82 (CI: 0.65 - 1.04)) and at maximum follow-up (adjusted relative hazard 0.80 (CI: 0.66 - 0.96)).
Clinical conclusions Patients of cardiologists were more likely to undergo right-heart catheterisation or cardiac catheterisation and had higher odds for transfer to an intensive care unit and electrocardiographic monitoring. This study found no survival advantage attributable to specialty care in the short period during which it was most likely to have made a difference. A trend was observed toward improved survival among patients of cardiologists at 1 year and at maximum follow-up, but the authors could not explain this relative difference in mortality with the available data.
Measure of benefits used in the economic analysis The authors did not provide a summary measure of benefit in the economic analysis. As such, a cost-consequences analysis was performed and the health benefits were associated with the health outcomes reported above.
Direct costs Costs were not discounted. Quantities and costs were reported separately. Hospital charges were considered, these being obtained from the hospital billing system. They were then used to estimate costs by using Medicare cost-to-charge ratio for each cost centre (Medicare uniform bill, 1982 version) and were converted to 1994 dollars by using the Consumer Price Index.
Statistical analysis of costs Wilcoxon rank-sum test was used for bivariable comparisons. Linear regression modelling was used in analyses of costs adjusting for the effects of confounding variables.
Sensitivity analysis The validity of findings was tested by using involved multiple subset and stratified analyses. The sensitivity of findings to an unmeasured confounder was assessed using the method described by Lin et al (1998) which calculates how unevenly an unmeasured confounder would need to be distributed between patients of cardiologists and patients of generalists in order to change the effect of physician speciality.
Estimated benefits used in the economic analysis As no summary benefit measure was provided, the reader is referred to the effectiveness results reported above.
Cost results The median unadjusted 1994 cost was $11,800 for patients of cardiologists and $5,000 for patients of generalists, (p=0.001). The median cost increment attributable to cardiologist care after adjustment was $2,100- 42.9% (95% CI: 27.8% - 59.8%).
Synthesis of costs and benefits Not applicable due to the cost-consequences approach adopted.
Authors' conclusions Cardiologists' care was associated with greater costs and resource use and no difference in survival at 30 days of follow-up. Whether the trend toward better survival at longer follow-up represents differences in care or unadjusted illness severity is uncertain.
CRD COMMENTARY - Selection of comparators No specific health technology was regarded as the comparator.
Validity of estimate of measure of benefit The effectiveness results are likely to be internally valid given the prospective study design and adjustments made for the effects of the confounders, addressing lack of patient comparability in baseline characteristics between the study groups. The study may be regarded as a cost-consequences analysis.
Validity of estimate of costs Quantities were reported separately from the costs and adequate details of methods of cost estimation were given. Cost results may not be generalisable to other settings.
Other issues The authors' conclusion seems to be justified given the adjustments made for confounders and the sensitivity analyses performed. The authors acknowledged that SUPPORT patients represented a severely ill, hospitalised population with poor prognosis, and the results may not be generalisable to other patients with congestive heart failure. Appropriate comparisons were made with other studies.
Implications of the study The implications of cost and survival differences among patients with congestive heart failure managed by cardiologists and generalists require further study to understand how to optimise care for these patients.
Source of funding Grant support from the Robert Wood Johnson Foundation.
Bibliographic details Auerbach A D, Hamel M B, Davis R B, Connors A F, Regueiro C, Desbiens N, Goldman L, Califf R M, Dawson N V, Wenger N, Vidaillet H, Phillips R S. Resource use and survival for patients hospitalized with congestive heart failure: differences in care by specialty of the attending physician. Annals of Internal Medicine 2000; 132(3): 191-200 Other publications of related interest Comment in: Annals of Internal Medicine 2000;132(3):238-9.
The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT). The SUPPORT Principal Investigators. A controlled trial to improve care for seriously ill hospitilized patients. Journal of the American Medical Association 1995;274:1591-8.
Lynn J, Knaus W A. Background for SUPPORT. Journal of Clinical Epidemiology 1990;43:1S-4S.
Kreling B, Robinson D K, Bergner M. Data collection strategies in SUPPORT. Journal of Clinical Epidemiology 1990;43:5S-9S.
Murphy D J, Knaus W A, Lynn J. Study population in SUPPORT: patients (as defined by disease categories and mortality projections), surrogates, and physicians. Journal of Clinical Epidemiology 1990;43:11S-28S.
Lin D Y, Psaty B M Kronmal R A. Assessing the sensitivity of regression results to unmeasured confounders in observational studies. Biometrics 1998;54:948-63.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Age Factors; Aged; Cardiology; Cost-Benefit Analysis; Family Practice; Female; Health Resources /economics /utilization; Heart Failure /mortality /therapy; Hospitalization /economics; Hospitals, Teaching; Humans; Male; Middle Aged; Outcome Assessment (Health Care); Practice Patterns, Physicians'; Prospective Studies; Referral and Consultation; Sensitivity and Specificity; Severity of Illness Index; Statistics as Topic; United States AccessionNumber 22000008026 Date bibliographic record published 31/07/2000 Date abstract record published 31/07/2000 |
|
|
|