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.

Several diagnostic strategies for patients with acute, undifferentiated chest pain were examined.

Strategy 0 was to discharge all patients without additional testing.

Strategy 1 was cardiac enzyme testing at presentation.

Strategy 2 was observation until at least 6 hours after onset of pain then cardiac enzyme testing.

Strategy 3, as for strategy 2, but followed by an exercise stress test (EST) if the blood tests were negative.

Strategy 4 was to admit to hospital for 24 hours then cardiac enzyme testing (no EST before discharge).

Strategy 5, as for strategy 4, but followed by an EST if the blood tests were negative.

Diagnosis.

Cost-utility analysis.

The study population comprised a hypothetical cohort of 1,000 patients presenting to hospital with acute chest pain unexplained by trauma or chest radiological finding, and:

no electrocardiogram changes diagnostic of acute myocardial infarction (AMI) or unstable angina (UA);

either risk factors for coronary heart disease (CHD) or clinical features suggesting CHD (i.e. patients at negligible risk of CHD were excluded);

no evidence of other serious abnormality requiring hospital admission; and

no clinically obvious UA, defined as known CHD with prolonged or recurrent episodes of cardiac type chest pain.

The setting was secondary care. The economic study was carried out in the UK.

The effectiveness data were derived from studies dating from 1982 to 2002. The price year was 2000/2001.

The evidence for the final outcomes was derived from a review of published studies and authors' assumptions.

A decision analytical model with six branches, each representing a diagnostic strategy was created using DATA decision analysis software (version 3.5; TreeAge Software). Strategies were eliminated if they were dominated (i.e. if another strategy saved more QALYs for the same or lower cost). They were also eliminated if they were subject to extended dominance by a combination of two other strategies (i.e. their incremental cost-effectiveness compared with the cheaper strategy exceeded the incremental cost-effectiveness of the more expensive strategy).

The outcomes assessed were:

the prevalence of AMI and UA;

the 30-day mortality of AMI and UA;

the sensitivity of each of the six strategies for AMI and UA;

the specificity of cardiac enzyme tests for each of the six strategies;

the 30-day mortality of untreated AMI and UA;

the treatment effect upon short-term mortality of untreated AMI, UA and non-cardiac pain (NCP); and

the treatment effect upon the infarction rate of UA.

Not reported.

Not reported.

Not reported.

Not reported.

Approximately 23 studies were included in the review.

Not reported.

Not reported.

The prevalence of AMI was 4% and that of UA was 8%.

The sensitivity for AMI was 0 with strategy 0, 0.45 with strategy 1, 0.85 with strategies 2 and 3, and 0.98 with strategies 4 and 5.

The sensitivity for UA was 0 with strategy 0, 0.10 with strategy 1, 0.20 with strategy 2, 0.75 with strategies 3 and 5, and 0.50 with strategy 4.

The specificity of cardiac enzyme tests was 1.00 for strategy 0, 0.95 for strategies 1, 2 and 4, and 0.75 for strategies 3 and 5.

The 30-day mortality of untreated AMI was 8% and that of untreated UA was 6%.

The infarction rate for UA was 10%.

The treatment effect upon short-term mortality of untreated AMI, UA and NCP was 0.5.

The treatment effect upon the infarction rate of UA was 0.5.

The authors made some assumptions when deriving estimates of effectiveness, to supplement those obtained from the review of the literature. To obtain estimates of survival, the authors made assumptions based on results from the literature.

The authors assumed that short-term mortality was zero for NCP regardless of any treatment given.

Patients with NCP were assumed to have the same annual mortality rate as the age-adjusted, normal UK population.

For patients with UA and AMI, survival over the first 3 years was estimated using data from a multi-centre chest pain study. Following this, the annual mortality was assumed to be 3.5 times the age- and gender-adjusted rate. Values of 23.24 years for NCP, 12.68 years for AMI and 12.96 years for UA were obtained. Discounting life expectancies at a rate of 6% per annum yielded 13.08 years for NCP, 9.2 years for AMI and 9.33 years for UA.

The measure of benefits was the QALYs saved. Health utility data were derived from a study of patients attending the Northern General Hospital (NGH) Chest Pain Observation Unit (CPOU). The discounted quality-adjusted life expectancy was 9.28 for NCP, 5.98 for AMI and 6.06 for UA.

The direct costs included in the study were those of the health service. These costs were for medical admission, medical treatment, cardiac enzyme tests, treating AMI and UA, terminal care and investigating false-positive tests. The direct costs of running each strategy were then estimated by summing its constituent elements. The costs of the initial Accident & Emergency department assessment were omitted from the analysis, as they were assumed to be the same in all cases. The unit costs for medical admission were obtained from published estimates, while those for medical treatment were obtained from the British National Formulary. The costs of cardiac enzyme tests were estimated using data from the NGH pathology department. The costs for treating AMI and UA were estimated using data from patients admitted after assessment in the NGH CPOU. The costs of investigating false positives were estimated using data for patients with NCP admitted after CPOU evaluation. The cost of terminal care was estimated from published unit costs of death in the Accident & Emergency department. Discounting was relevant since all the costs were incurred during 13 years. Hence, the costs were appropriately discounted at a rate of 6% per annum. The study reported the total and incremental costs per 1,000 patients. The price year was 2000/2001.

The costs were treated as point estimates (i.e. the data were deterministic).

The indirect costs were not included in the analysis.

UK pounds sterling ().

One-way sensitivity analyses were undertaken for costs, utilities, treatment effects, mortality and infarction rates, and prevalence. Each parameter was varied through a range of values to determine how much this changed the findings of the analysis. The assumption that patient utility with NCP is unaffected by diagnosis was also tested in the analysis.

The expected discounted QALYs per 1,000 patients were 8,853.7 for strategy 0, 8,859.4 for strategy 1, 8,864.7 for strategy 2, 8,870.2 for strategy 4, 8,872.5 for strategy 3, and 8,873.8 for strategy 5.

Therefore, the incremental discounted QALYs per 1,000 patients were 5.7 for strategy 1 over strategy 0, 5.3 for strategy 2 over strategy 1, 5.5 for strategy 4 over strategy 2, 2.3 for strategy 3 over strategy 4, and 1.2 for strategy 5 over strategy 3.

As strategy 4 was subject to extended dominance, and hence was excluded from the analysis, the incremental discounted QALYs gained per 1,000 patients by using strategy 3 over strategy 2 was 7.8.

The total discounted costs per 1,000 patients were 1,399,700 for strategy 0, 1,499,600 for strategy 1, 1,597,100 for strategy 2, 1,796,100 for strategy 4, 1,820,800 for strategy 3, and 1,970,500 for strategy 5.

Thus, the total incremental discounted costs per 1,000 patients were 99,900 for strategy 1 over strategy 0, 97,500 for strategy 2 over strategy 1, 199,000 for strategy 4 over strategy 2, 24,700 for strategy 3 over strategy 4, and 149,700 for strategy 5 over strategy 4.

As strategy 4 was subject to extended dominance, and hence was excluded from the analysis, the incremental discounted costs per 1,000 patients of using strategy 3 over strategy 2 were 223,700.

The costs and benefits were combined by calculating a cost-utility ratio (the additional cost required per QALY gained). The incremental cost per QALY was 17,432 when strategy 1 was used over strategy 0, 18,567 when strategy 2 was used over strategy 1, 28,553 when strategy 3 was used over strategy 2, and 120,369 when strategy 5 was used over strategy 3.

The estimates of the incremental analysis were comparatively insensitive to variations in the prevalence of AMI or UA, the utilities of AMI or UA, mortality estimates and treatment effect estimates. They were also comparatively insensitive to the costs of treating AMI or UA, the costs of terminal care, and the costs of long-term treatment of survivors. However, the results were sensitive to variations in the cost of each strategy, the cost of ruling out cardiac disease, and the effect of diagnostic testing upon the utility of NCP.

The diagnostic strategies for acute, undifferentiated chest pain that entailed observation and cardiac enzyme testing, with or without exercise stress testing, had similar cost-effectiveness ratios to currently funded cardiovascular interventions. However, strategies requiring overnight admission for all patients appeared to be poor value for money.

The authors compared six different diagnostic strategies with each other. The six different strategies appear to have represented all the potential diagnostic strategies currently possible. You should decide if these are widely used diagnostic strategies in your own setting.

Even though the authors did not report that a systematic review of the literature had been undertaken, they appear to have derived their estimates of effectiveness from an exhaustive review of approximately 30 studies. Hence, it would appear that relevant research was identified and biases were minimised. Due to constraints imposed by the journal in which the paper was published, the authors were only able to give brief descriptions of the data sources. However, those readers requiring further details were encouraged to contact the authors. Due to these constraints, it remains unclear how the results from different studies were combined, and how any differences between the studies were considered. However, the authors did provide a range of values next to the baseline estimates, which were varied in the sensitivity analysis. To derive effectiveness data with which to supplement the data from the review of the literature, the authors made assumptions, many of which were based on the literature. All of their assumptions were appropriately tested in the sensitivity analysis.

The estimation of benefits was modelled. The instrument used to derive a measure of health benefit was appropriate. All the benefits were appropriately discounted at a rate of 6% per annum.

All the categories of cost relevant to the health service perspective adopted appear to have been included in the review and, as such, all the relevant costs have been included. The costs of the Accident & Emergency department were excluded as they were common to all six diagnostic strategies. The authors did not include the costs of litigation, arguing that it would represent double counting of the costs and disbenefits. The costs and the quantities were not reported separately because the costs were derived through modelling. All the costs were appropriately varied in the sensitivity analysis, using ranges that appear to have been appropriate. The costs were incurred during a 13-year period and discounting was, appropriately, performed. The price date was reported, which will ease any possible inflation exercises.

The authors did not compare their results with those from other studies. The issue of generalisability to other settings was addressed through the sensitivity analysis. The authors do not appear to have presented their results selectively and their conclusions reflected the scope of the analysis. The authors reported a further limitation to their study. They assumed that all patients in the population would undergo the strategy provided, whereas in practice, patients may be selected on the basis of their clinical features or by the clinicians' decision.

The authors concluded that the combination of two to six hours of observation and exercise stress testing appears to be effective and reasonably cost-effective. However, they recommended that this strategy should be carefully evaluated in practice before widespread implementation is considered, as more information about the effect of test results upon individuals with NCP is needed.

None stated.

Herren KR, Mackway-Jones K, Richards CR, et al. Diagnostic cohort study. Is it possible to exclude a diagnosis of myocardial damage within six hours of admission to an emergency department? BMJ 2001;323:372-5.

Haim M, Gottlieb S, Boyko V, et al. Prognosis of patients with a first non-Q wave myocardial infarction before and in the reperfusion era. American Heart Journal 1998;136:245-51.

Hamm CW, Goldman BW, Heeschen C, et al. Emergency room triage of patients with acute chest pain by means of rapid testing for cardiac troponin T or troponin I. New England Journal of Medicine 1997;337:1648-53.