Table S1.

The correlation between MUR-PSQ-v2 items

Item no	1	2	3	4	5	6	7	8	9	10	11	12
1	1.00	0.433	0.588	0.530	0.530	0.547	0.418	0.497	0.500	0.524	0.512	0.582
2	0.433	1.00	0.526	0.458	0.429	0.399	0.299	0.402	0.453	0.410	0.430	0.470
3	0.588	0.526	1.00	0.714	0.658	0.655	0.346	0.617	0.535	0.572	0.565	0.652
4	0.530	0.458	0.714	1.00	0.634	0.642	0.298	0.606	0.457	0.502	0.521	0.599
5	0.530	0.429	0.658	0.634	1.00	0.716	0.341	0.684	0.507	0.492	0.548	0.602
6	0.547	0.399	0.655	0.642	0.716	1.00	0.425	0.685	0.576	0.611	0.625	0.677
7	0.418	0.299	0.346	0.298	0.341	0.425	1.00	0.463	0.567	0.555	0.574	0.468
8	0.497	0.402	0.617	0.606	0.684	0.685	0.463	1.00	0.647	0.624	0.609	0.669
9	0.500	0.453	0.535	0.457	0.507	0.576	0.567	0.647	1.00	0.715	0.670	0.680
10	0.524	0.410	0.572	0.502	0.492	0.611	0.555	0.624	0.715	1.00	0.825	0.757
11	0.512	0.430	0.565	0.521	0.548	0.625	0.574	0.609	0.670	0.825	1.00	0.756
12	0.582	0.470	0.652	0.599	0.602	0.677	0.468	0.669	0.680	0.757	0.756	1.00

Notes: The correlation between the variables was examined by creating a correlation matrix. Because the MUR-PSQ-v2 was constructed to measure one concept (patient satisfaction with an MUR), there is an expectation that the variables will correlate with each other. However, any variable with a consistently low correlation (less than 0.2) with other variables is potentially redundant. Equally, a variable with a consistently high correlation (over 0.8) would cause multilinearity problems during factor analysis and might also need removing.

Abbreviation: MUR-PSQ-v2, Medicines Use Review patient satisfaction questionnaire, version 2.