Instruments to measure shared decision-making in outpatient chronic care: a systematic review and appraisal

1. Background

Patient involvement during chronic disease care improves outcomes [1]. One approach to foster involvement is mutual patient-provider decision-making about a plan of care termed, “shared decision-making” (SDM) [2]. Initiatives including Evidence Communication Innovation Collaborative have promoted SDM prompting a closer look at how decisions are jointly made [3]. Still, the rate at which providers involve patients in decision-making remains low [4]. Despite increasing demands for SDM, effectiveness in chronic disease care has not been well described [5,6], likely because no “gold standard” measuring SDM in research and practice exists [7].

While SDM instruments have been tested, few undergo comprehensive psychometric evaluation thus questioning how robustly the instrument performs. Rigorous appraisals of SDM instruments are warranted. The aim of this study was to (1) evaluate reliability and validity of existing instruments; (2) assess each instrument’s ability to capture SDM elements.

2. Theoretical underpinnings

Makoul et al. [8] proposes that SDM is composed of 23 elements, nine of which are essential and 14 elements considered ideal (e.g., presenting evidence) and/or general (e.g., partnership).

3. Methods

A comprehensive search was conducted across five electronic databases using PRISMA [9] guidelines. We also searched for gray literature (abstracts, conference proceedings, and unpublished manuscripts). Inclusion criteria stipulated that manuscripts be English language reports of SDM instrument development, with testing, and validation in ambulatory chronic disease care.

Three researchers performed psychometric appraisal using Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN) checklist [10] to evaluate nine measurement properties: internal consistency, reliability, measurement error, content validity, construct validity, criterion validity, and responsiveness. The researchers next evaluated each instrument to determine the presence of SDM elements.

4. Results

Sixteen instruments across 30 studies were eligible for appraisal (See Table 1).

Table 1.

Comparison of shared decision-making instruments by measurement properties and integrative model elements

Instrument characteristics	CollaboRATE [11–13]	COMRADE [14]	DEEP-SDM [15]	Individual care [16]	Informed decision-making tool [17]	MAPPIN’SDM [18,19]	OPTION5 [20–22]	OPTION/OPTION revised [7,23,24]	OPTION (dyadic) [25,26]	OPTION12 [21,22,27]	PACIC [28]	PICS [29]	SDM MASS [30]	SDM-Q [31]	SDM-Q-9 [11,32–34]	SDM Q-Doc [35]
Instrument perspective	P	P	O	P	O	C P O	O	O	C P	O	P	P	C P O	P	P	C
Number of items	3	20	-	37	9	11	5	18	12	12	20	25	15	15	9	9
Internal consistency reliability (Cronbach’s alpha)	0.9			0.96		0.87	0.79	0.68	0.9	0.68	0.94	0.73	0.87	0.77	0.94	0.91
Construct validity		√		√				√		√	√	√			√	√
Makoul’s integrative SDM model [8]
1.Essential SDM elements
Define/explain problem	√	√	√		√	√	√	√	√	√		√	√	√	√	√
Present options		√	√		√	√	√	√	√	√		√	√	√	√	√
Discuss pros/cons		√	√		√	√	√	√	√	√		√	√	√	√	√
Patient values/preferences	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√
Discuss patient ability/self-efficacy		√	√	√	√	√	√	√	√	√	√	√	√	√	√	√
Doctor knowledge/recommendations		√	√		√	√	√	√	√	√	√	√	√	√	√	√
Check/clarify understanding	√	√	√	√	√	√	√	√	√	√		√	√	√	√	√
Make or explicitly defer decision		√			√	√	√	√	√	√			√	√	√	√
Arrange follow-up	√		√			√		√	√	√	√		√	√	√	√
2. Ideal SDM elements
Unbiased information		√	√				√				√
Define roles			√	√	√	√	√	√	√		√	√	√	√	√	√
Present evidence	√	√	√		√	√	√	√	√	√	√	√	√	√	√	√
Mutual agreement		√			√		√	√	√	√		√		√	√	√
3. General SDM qualities
Deliberation/negotiation				√	√		√	√	√	√		√		√	√	√
Flexibility/individualized approach	√	√	√		√		√	√	√	√	√	√		√	√	√
Information exchange	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√
Involves at least two people	√	√	√		√	√	√	√	√	√		√	√	√	√	√
Middle ground		√					√	√		√		√		√	√	√
Mutual respect							√	√	√	√		√		√	√	√
Partnership		√	√		√		√	√		√		√		√	√	√
Patient education		√	√	√	√	√	√	√	√	√	√	√	√	√	√	√
Patient participation	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√	√
Process/stages	√		√				√	√	√	√	√	√		√	√	√
Total elements	9	18	18	8	18	15	22	22	20	21	12	20	15	22	22	22

Abbreviations: SDM, shared decision-making; C, clinician reported; P, patient reported; O, observer reported; COSMIN, Consensus-based Standards for the Selection of Health Measurement Instruments.

Reprinted with permission [8].

Instruments measured SDM from different perspectives: provider (n = 4), patient (n = 11), and/or observer (n = 7). Internal consistency reliability was highest among individual care [16] (r = .96), SDM-Q-9 [32,33] (r = .94), and MAPPIN’SDM [18,19] (r = .94). Most studies did not describe the handling of missing data thus threatening internal consistency. Concurrent and convergent validity was tested in 12 instruments. Construct validity was reported for eight instruments. Goodness of fit and differential item functioning was rarely reported.

Overall, SDM-Q-9 [32,33] and the adaptations of OPTION [20,23,25,27] included the most elements. Seven of the sixteen instruments met 100% essential SDM elements. Five instruments missed “Arrangement of follow-up”. Only OPTION5 [20] was found to have all 4 “ideal” elements.

5. Discussion

Our review suggests that SDM-Q-9 [32,33] and the adaptations of OPTION [20,23,25,27] are superior to other instruments measuring SDM evidenced by high reliability, validity, and presence of SDM elements. Because SDM involves a set of reciprocal behaviors, it is unclear whether instruments from the perspective of an observer are reliable given low correlations between observer assessments of SDM and patient reports [4].

6. Conclusion

This study evaluated reliability, validity, and inclusion of SDM elements across 16 instruments. It appears SDM-Q-9 [32,33] and versions of OPTION [20,23,25,27] have the most robust psychometric testing and were most inclusive of items capturing SDM elements. Further testing is needed to establish the acceptability and feasibility of instrument use in clinical and research settings.

What is new?

Key findings

• Our findings determined that SDM-Q-9 and the various adaptations of OPTION appear to be the most robustly tested instruments and include the most elements of shared decision making.

What this adds to what was known?

• After systematically evaluating 16 instruments, the measurement of shared decision making remains variable across studies and settings.

What is the implication and what should change now?

• Future research that further establishes the psychometric properties of instruments measuring SDM, in addition to testing the feasibility and acceptability, is warranted.