Development of key performance indicators to evaluate centralized intake for patients with osteoarthritis and rheumatoid arthritis

Abstract

IntroductionCentralized intake is integral to healthcare systems to support timely access to appropriate health services. The aim of this study was to develop key performance indicators (KPIs) to evaluate centralized intake systems for patients with osteoarthritis (OA) and rheumatoid arthritis (RA).MethodsPhase 1 involved stakeholder meetings including healthcare providers, managers, researchers and patients to obtain input on candidate KPIs, aligned along six quality dimensions: appropriateness, accessibility, acceptability, efficiency, effectiveness, and safety. Phase 2 involved literature reviews to ensure KPIs were based on best practices and harmonized with existing measures. Phase 3 involved a three-round, online modified Delphi panel to finalize the KPIs. The panel consisted of two rounds of rating and a round of online and in-person discussions. KPIs rated as valid and important (≥7 on a 9-point Likert scale) were included in the final set.ResultsTwenty-five KPIs identified and substantiated during Phases 1 and 2 were submitted to 27 panellists including healthcare providers, managers, researchers, and patients in Phase 3. After the in-person meeting, three KPIs were removed and six were suggested. The final set includes 9 OA KPIs, 10 RA KPIs and 9 relating to centralized intake processes for both conditions. All 28 KPIs were rated as valid and important.ConclusionsArthritis stakeholders have proposed 28 KPIs that should be used in quality improvement efforts when evaluating centralized intake for OA and RA. The KPIs measure five of the six dimensions of quality and are relevant to patients, practitioners and health systems.Electronic supplementary materialThe online version of this article (doi:10.1186/s13075-015-0843-7) contains supplementary material, which is available to authorized users.

A Systematic Review of Quality Measures for Inflammatory Arthritis

“…From the 13 sources ( 16,17 , and 1 was a pan-European set 18 ). Most of the sets were for RA 6,7,9,11,12,13,[15][16][17][18] (n = 10, 77%) with 1 each (8%) for JIA 10 , PsA 14 , and inflammatory arthritis 8 . No sets were found for ReA and none accepted for AS.…”

“…Similarly, there was variability in the consensus procedures used to finalize the measurement sets: a modified Delphi in conjunction with Nominal Group Technique in 1 set 10 , a modified RAND/University of California, Los Angeles (UCLA) Appropriateness Method in 4 sets 11,12,16,17 , an online modified Delphi using a platform called ExpertLens (RAND Corp.) in 3 sets 6,7,8 , a modified Delphi in another 17 , and other study-specific consensus methods in 4 sets 9,13,14,15 .…”

“…Interestingly, only a few of these attributes are routinely reported in the studies describing the measure development of the inflammatory arthritis measure sets in this review. For example, to be accepted, quality measures needed variably to be valid 11 , valid and feasible 7,12 , valid and important 6 , valid and relevant 8 , relevant 17 , and applicable and feasible 18 . A few sets did not follow any published attributes for measure development and used the following statements for consensus development: "adequately reflected the quality of disease course monitoring" 16 , "consistent with [previous] recommendations" 9 , or simply included measures with higher "priority" rankings 10,13,14,24 .…”

“…More recently, the RAND/UCLA method has been adapted for online panels using a platform called ExpertLens 21 . This platform was used for 3 panels 6,7,8 and has the advantage of including a larger number of participants (40-50) than traditional consensus panels 20 , allowing for broader input. While there is no current standard methodology for quality measure development, at a minimum it is recommended to develop measures based on a high level of evidence (usually from a guideline).…”

“…For example, considering the evidence base used for the measure development, only 3 sets of quality measures were based on systematic reviews of existing guidelines and measures 7,16,18 , with another 3 based on nonsystematic literature reviews 6,8,11 . The rest of the measures were developed based on discussions and/or questionnaires with those involved or on existing guidelines.…”