Association between procedure appropriateness and patientreported outcomes after percutaneous coronary intervention

Objective The Appropriate Use Criteria (AUC) has been used to identify individuals who are likely to benefit from percutaneous coronary intervention (PCI) for stable ischaemic heart disease. However, whether physicians reliably grade PCI appropriateness and whether AUC categories stratify symptomatic improvement in real-world practice are unclear. Methods Patient-reported outcomes measures (PROMs) for angina (Seattle Angina Questionnaire (SAQ-7)), dyspnoea (Rose Dyspnea Scale (RDS)) and depression (Patient Health Questionnaire-2 (PHQ-2)) were collected on patients undergoing elective coronary angiography at an academic medical centre. Retrospectively, two physicians independently determined PCI appropriateness by the AUC criteria. Results Inter-rater agreement on appropriateness was moderate (kappa=0.48, 95% CI 0.32 to 0.63). Of PCI procedures evaluated, 57 (47.1%) were appropriate (A-PCI), 62 (51.2%) were maybe appropriate (MA-PCI) and 2 (1.6%) were rarely appropriate. At baseline, A-PCI compared with MA-PCI patients had worse RDS scores (2.0 vs 1.2, p=0.01). At 30 days, the change in SAQ-7 summary score was similar between groups (A-PCI vs MA-PCI, +27.1 vs +20.7, p=0.11). The mean change in RDS score was greater in A-PCI than MA-PCI (-1.0 vs -0.5, p for group by time interaction=0.03). PHQ-2 scores were similar and did not improve at 30 days. Conclusion In patients undergoing PCI with PROMs collected before and 30 days after PCI, similar improvements in angina were observed regardless of appropriateness. Inter-rater agreement on PCI appropriateness was only moderate. Use of PROMs may improve reliability of physician assessments of PCI appropriateness.