Anything that can be measured can be improved: the case of performance measures in heart failure and beyond

This editorial refers to ‘Association between process performance measures and 1-year mortality among patients with incident heart failure: a Danish nationwide study’, by A. Nakano et al., on page 28. “Anything that can be measured can be improved.” (Michael Dell) There is a longstanding interest in the development and application of quality of care metrics in order to improve quality of health care and subsequently patient outcomes. Performance measures quantify processes, structures, efficiencies or outcomes and are usually based on strong guideline recommendations that are associated with clinically meaningful patient outcomes. Performance measures must be clearly defined, reliable, valid, feasible, and actionable. They partially cover the continuum of care; i.e. diagnosis and treatment, patient education, and self-management. Their use is intended for public reporting, comparisons between institution and/or healthcare providers and pay-for-performance programmes.1–3 Nakano et al.4 present a Danish nationwide study on the association of performance measures and mortality in 24 308 patients with incident heart failure, including 17 758 patients with a left ventricular ejection fraction (LVEF) ≤ 40%. The authors defined six performance measures; namely echocardiography and New York Heart Association (NYHA) classification (diagnosis), appropriate medication with angiotensin-converting enzyme (ACE)/angiotensin II (ATII) inhibitors and beta blockers (treatment), patient education, and physical training (education and self-management), thereby covering different domains of care. Overall, quality of care was expressed as ratio of fulfilled to total performance measures thereby assuming equal importance of performance measures. In other words, documentation of NYHA class was equally weighted as ACE/ATII inhibitor use which is at least questionable. By making use of different well-known Danish registries, the authors analysed the degree of guideline adherence for the aforementioned processes and its relation with mortality. Within 1 year, 17.1% of patients had died. Except for beta blocker use and physical training, implementation of the remaining four performance measures was associated with reduced mortality. The aggregate performance measure was also associated with mortality. For instance and of utmost importance; if >50 or >75% of processes were fulfilled, mortality was reduced by an astonishing 30 and 60%, respectively, compared with patients with <25% process fulfilment. While in the latter category mortality was about 20%, it was about 14% and 7% in the former groups. Fulfilment of performance measures ranged from 9% (physical training) to 79% (echocardiography). In patients with LVEF <40%, only 70% and 63% of patients had ACE/ATII inhibitors and beta blockers, respectively. Considering this moderate use and the large impact on outcome, there seems to be substantial room for improvement in heart failure care. The authors can be congratulated for this study which is an important contribution to better understand the association between quality of care and outcomes that must now be translated in quality improvement programmes. From our perspective, there are two main clinically relevant take away messages; (i) heart failure care is suboptimal at best and (ii) there is a dose–response association between guideline-adherent heart failure management and mortality. So what are unresolved issues and practical implications? First, the authors selected certain heart failure performance measures, while not evaluating others such as weight and blood pressure measurement or smoking cessation counselling.1 The appropriate use of cardiac rhythm management devices such as implantable defibrillators or cardiac resynchronization therapy was not addressed at all although being part of other suggested performance measure sets.2 In addition, this study did not report atrial fibrillation (AF) as co-morbidity which is known to occur in up to 50% of the heart failure population. On the one hand, AF also requires additional performance measures such as oral anticoagulation use.3 On the other hand, therapies; especially beta blockers may reduce mortality in heart failure patients with reduced LVEF and sinus rhythm but not in AF.5 Finally, this study focused on mortality, while not evaluating other important outcome measures such as rehospitalization or quality of life which was beyond the scope of this investigation. In conclusion, we believe it is time to measure; it is time to improve. In our view, this should be done on a large scale with standardized instruments and outcome measures. For instance, we have recently initiated a novel AF quality management program at Helios hospitals. By making use of claims data of the entire hospital network, we regularly monitor mortality in AF6 and also severe complications of AF catheter ablation (Figure 1).7 This is complemented by a case-based internal peer review process providing actionable quality improvement suggestions. Other related initiatives are the development and implementation of standardized patient-reported outcome measures and large scale blended learning programs for guideline-adherent management in heart failure and cardiac arrhythmias. Helios hospital network atrial fibrillation quality management programme. In-hospital mortality rates for atrial arrhythmias (left panel, atrial fibrillation and atrial flutter) and left and right atrial catheter ablation (right panel) are shown. Results are based on the analysis of claims data of 21 391 (Helios hospitals) and 303 787 cases (Germany) in 2016. Courtesy of Dr. König, Leipzig. Hopefully, the Nakano study,4 our activities6,7 and those of many others in this field will further increase the interest, awareness and implementation of quality management programmes in heart failure and beyond with the ultimate goal of better and longer patient lives. Conflict of interest: none declared.