Indicators of quality of care for patients with acute myocardial infarction

Nov. 10th, 2008

Dear Dr. Tu:

Your excellent review on the "Indicators of quality of care for patients with acute myocardial infarction" published in the CMAJ October issue, 2008, is very informative and likely to be instrumental in improving the over all quality of care of patients suffering from the condition. Your panel must be congratulated to have made the landmark studies.

Some aspects on the practical assessment of adherence to the indicators may merit discussion and certain suggestions ventured to be made: -

A.) The setting of target benchmark at 90% of the indicators: - This may sometimes be less than ideal, as every indicator is considered having an equal unit value. The indicators are not "weighed" (given variable points) according to their relative importance. By fulfilling most of the non-urgent, non-lifesaving indicators yet failing to meet some urgent, lifesaving indicators, one may still attain the crude 90% benchmark, thus giving the false impression of having provided a good quality of care. Conversely, if most or all of the more important indicators have been met, leaving some of the less significant indicators unfulfilled, one actually is providing a good quality of care but the benchmark attained may be ranked below the 90% target level. Meeting or failing to meet the selected target benchmark percentage therefore does not equate to good care or bad care. At best, the target benchmark may help the data extraction staffs to flag a fraction of the cases with the need for in depth analysis by physicians. The quality of care of those cases only can be meaningfully assessed.

B.) The idea of "weighed" indicators: - Having identified a series of valid indicators as you had so aptly done, the question arises as to the merit of categorizing and weighing each indicator for the purpose of calculating the more meaningful target benchmark percentages. It may be suggested that the indicators be categorized as a) Critical, urgent and potentially immediately life saving (e.g. primary PCI within 90 minutes after hospital arrival etc); b) Critical but may be delayed as long as fulfilled within the time frame of the current hospitalization; potentially improving quality of life or reducing hospital complications (e.g. assessment of left ventricular function etc.) and c) Elective, optional may even be carried out in the outpatient follow-up clinic; potentially improving long term outcome or preventing recurrences (e.g. statin prescription on hospital discharge; cessation of smoking etc.). Each indicator may be assigned a numerical value ranging from 1 to 3 points as acceptable to the panel members. For example 3 points are assigned to each indicator of category a); 2 point to each of category b) and 1 point to each of category c). This will greatly facilitate the accurate flagging of cases for in depth analysis. By virtue of proper selection of indicators, and the probability of the sum total of points from categories a) & b) will normally exceed those from category c), the following statements will be valid. With weighing of the indicators it literally requires the fulfillment of a great majority of categories a) & b) indicators before any one can attain the 90% benchmarks. Conversely, by fulfilling most of the category c) indicators without largely adhering to the category a) & b) indicators one will invariably attain a benchmark level significantly below 90%.

C.) The cut-off percentage benchmarks for cases to be analyzed by physicians in estimating the true quality of care: - Obviously in any series of cases, variable levels of benchmark percentages will be attained. The next question will be: - "Cases with which cut-off points should be chosen for in depth analysis without missing out any cases of substandard care?" It may be suggested that if categorization and assignment of points to the indicators denoted in your paper are made, two standard parameters may be calculated; namely the "Crude minimal performance index*" and the "Critical performance index**" (see footnotes). These two indices can then be used by the data extraction staff as standards for comparison with their respective actual Case performance index (crude)# and Case performance index (critical)## (see footnotes) attained in each of the cases of the series. Cases with their actual indices below either or both of the two corresponding standard indices will be flagged for physician assessment to determine whether adequate quality of care has been provided. Cases of substandard care will not be left undetected. Sequential comparison of different series of cases in the same hospital or series of cases from different hospitals may more accurately reflect changing trends in the quality of care. As early as in the 1980s the above principle and methodology of assessing quality of care was acknowledged by the American College of Utilization Review Physicians, incorporating the original publication among the prescribed studies for candidates taking their board examinations.

Hopefully the above suggestions may be considered in making your endeavours to improve the quality of care of patients with acute MI even more effective.

Yours truly,

Dr. W. T. WONG FRCS(England), FRCSC. Victoria, BC

Footnotes: -

"Standard" Performance Indices calculated from chosen Indicators (To be used as Standards for comparison): -

*Crude minimal performance index (Standard) is Sum total of indicator points from category a) & b) indicators / Sum total of indicator points available in all 3 categories of indicators (expressed as a percentage).

** Critical performance index (Standard) is Panel chosen target benchmark percentage x sum total of indicator points from category a) & b) indicators / Sum total of indicator points available in all 3 categories of indicators. (Expressed as a percentage). (As indicated in your paper the panel may vary the chosen target benchmark percentage with time in order to foster progressive higher standards of care)

"Actual" Case Performance Indices computed on individual cases: -

#(Actual) Case performance index (crude) is Total indicator points scored in the case / Total indicator points available for all 3 categories a, b & c indicators (expressed as a percentage). If this is below the crude minimal performance index (standard), the case will be flagged for in depth review by physician.

##(Actual) Case performance index (critical) is Total indicator points scored in the case from categories a & b indicators / Total indicator points available for all 3 categories of indicators (expressed as a percentage). If this is below the Critical performance index (standard), the case will be flagged for in depth review by physician.

(The back slash "/" in the above formulae denotes "divided by")

Reference:

Wong W.T. Quality Assurance in Surgical Practice Through Auditing. The American Journal of Surgery Vol. 139 May 1980; 669-672

Conflict of Interest:

None declared

We respectfully disagree with several of the statements in this letter. Multiple ‘evidence-based’ meta-analyses of the randomized controlled trials of lipid-lowering therapy in patients who have recently suffered an acute coronary syndrome (ACS) have confirmed that statins reduce all-cause mortality in this high-risk group of patients.{1-4} Furthermore, they have demonstrated that benefits from statins are observed in the elderly, in women, and are a class-effect (including Atorvastatin) as opposed to being limited to a specific statin.{1-4} None of these analyses have concluded that statins are contraindicated in patients with congestive heart failure post-ACS.

References

1. Afilalo J, Majdan AA, Eisenberg MJ. Intensive statin therapy in acute coronary syndromes and stable coronary heart disease: a comparative meta-analysis of randomized controlled trials. Heart 2007; 93:914-21.

2. Josan K, Majumdar SR, McAlister FA. The efficacy and safety of intensive statin therapy: a meta-analysis of randomized trials. CMAJ 2008;278:576-84.

3. Afilalo J, Duque G, Steele R, et al. Statins for secondary prevention in elderly patients. J Am Coll Cardiol 2008;51:37-45.

4. Hulten E, Jackson JL, Douglas K, et al. The effect of early, intensive statin therapy on acute coronary syndrome. Arch Intern Med 2006; 166:1814-21.

Conflict of Interest:

None declared