Characteristics of included studies regarding study design, patient preference tool, outcome measures and relevance to multimorbidity
| Study ID (reference) | Type of study | Location | Target population and sample size | Setting of study | Name and type of tool used to elicit patient priorities and preferences | Outcome measured | Patient relevance of outcome | Validity testing | Incorporation into electronic medical records | Adverse effects |
|---|---|---|---|---|---|---|---|---|---|---|
| Asenlof (2009) | Prospective cohort | Uppsala, Sweden | Patients with subacute or chronic (>4weeks) musculoskeletal pain (n=54) | Primary care | Patient Goal Priority Questionnaire (PGPQ) and Patient Goal Priority List (PGPL) (Questionnaire) | Patient behavioural goals in the context of physical therapy. PGPL differs from PGPQ in that PGPL is a questionnaire that does not place restrictions on activities and is administered with the help of the physical therapist. | Not relevant, while the tool used was for measuring patient priority; it was used as a decision aid as opposed to a tool which was directly tied to patient health outcomes. | 52% raw percentage of agreement of the PGPQ questionnaire on two occasions. 14% raw percentage agreement between the PGPQ and PGPL questionnaires. | No | None |
| Carvalho, et al23 | Cross-sectional | California, USA | Pregnant women attending a prenatal class (n=82) | Outpatient | Pre-CS survey (Questionnaire) | Patient's preference in regard to potential outcomes in a caesarean delivery, measured by both a ranking and relative value score methodology. | Not relevant; the tool was a measure of patient preference for anaesthesia but was not directly tied to patient health outcomes before or after surgery. | Ranking and relative outcome scores for each outcome were positively correlated (R2=0.6–0.8 for 9 outcome variables, P<0.001). | No | None |
| Dierckx, et al24 | Cross-sectional | Ghent, Belgium | Patients >18 years attending physical therapy (n=237) | Outpatient | Control Preference Scale (CPS) (Questionnaire) | To document the patient’s preference in the degree of control they wish to have in their level of care. Also measured agreement of perceived patient level of control between the therapist and the patient. The OPTION tool was used to measure the degree to which patient involvement occurred during the decision-making process. | Not relevant; qualitative study model which did not measure the effectiveness of SDM vs no SDM but rather was looking at agreement between therapist and patient perception of level of involvement in care. | The CPS questionnaire was based on a Dutch version of the instrument which was not validity tested. | No | None |
| Fraenkel, et al25 | Cross-sectional | Connecticut, USA | Women with diagnosed systemic lupus erythematosus (n=103) | Primary care | Adaptive Conjoint Analysis (ACA) (Questionnaire) | Ask patients to choose between competing products, stimulating the way the patient would actually make choices faced with a task. Based on three assumptions: (1) Each treatment option can be broken down into attributes which are defined by a number of levels (eg, Risk of infection on medication A, B, and C are respectively 10%, 20% and 30%. (2) Respondents have unique values for each attribute level. (3) Utilities and values can be combined across attributes (eg, if the sum of the attributes for medication A is more than for medication B, then the patient should prefer medication A) | Not relevant; the goal of the study was to assess the effectiveness of adaptive conjoint analysis and also to see patient-related preferences on cytotoxic medications. The study did not measure changes in health outcome as a direct consequence of applying adaptive conjoint analysis. | Study did not explicitly measure the validity of the tool but instead reported good ‘face validity’. | No | None |
| Zolfaghari et al (2014) | Cross-sectional | Mannheim, Germany | Patients with atrial fibrillation on oral anti-coagulant therapy (n=180) | Primary care | Likert scale (Questionnaire) | The study looked to objectively identify the patient's preferences for the use of a vitamin K antagonist compared to direct oral anticoagulants. | Not a patient-relevant outcome measure, the tool was a measure of patient preference for anticoagulant therapy but did not measure direct changes to health outcome as a consequence of introducing the tool. | The reliability of the items measured in the questionnaire was tested and had a reliability of the seven items with a probability of 98%. | No | None;, did not present outcome results |
| Fried, et al26 | Cross-sectional | Connecticut, USA | Older patients (>65) with multiple chronic conditions (n=357) | Community dwelling | Priority (Questionnaire) | Measured patient priorities regarding the following outcome measures (Tool domains): “Keeping you alive maintaining independence pain management symptom management” | Not relevant, the study was validating health outcome prioritisation as a tool for decision-making. Study did not evaluate the effectiveness of such a tool on improving health outcomes but did use a tool relevant in the context of multimorbidity | Percentage agreement amount 35 participants who completed the tool regarding the most and least important outcome ranged from 85% to 100%. | No | None |