Assessment and determinants of patient satisfaction with mental health care

How satisfied patients are with the care they are receiving is widely regarded as an important process variable and quality indicator in mental health care.

It is a process variable, as it predicts to what extent the aim of care, i.e. the alleviation or overcoming of mental distress, may be achieved. Various studies have shown that more satisfied patients are more adherent to treatment and – even if there is no difference in adherence – benefit more from care than less satisfied patients. Furthermore, patient satisfaction predicts outcomes right from the initial stages of treatment, e.g. when assessed within the first two days of hospital care1. It is also a quality indicator, because all treatments should be as patient friendly as possible, independently of any impact on health and social outcomes.

Since the 1960s, numerous scales have been used to measure patient satisfaction with mental health care, also termed treatment satisfaction, service satisfaction or consumer satisfaction. A recent systematic review indicates that scales vary significantly in their structure, length, focus and quality2. There is no consensus on how exactly patient satisfaction should be measured and, across scales, patients are asked to rate their satisfaction with 19 different aspects of care. Despite an extensive literature, the review identified only four scales that have been used in more than 15 studies and may therefore be regarded as more established.

The Client Satisfaction Questionnaire3 for outpatient treatment and the Client Assessment of Treatment Scale4 for inpatient treatment are brief scales of 8 and 7 items respectively and provide global scores. The Verona Service Satisfaction Scale5 and the Self‐Rating Patient Satisfaction Questionnaire6 are much longer and have subscales on different care aspects in addition to a global score.

Satisfaction with care, as measured on such scales, can be influenced by characteristics of the patients and by aspects of the care they are receiving7.

A number of socio‐demographic characteristics such as gender, ethnicity, socio‐economic and marital status have been suggested as determinants of satisfaction with care, but the associations are usually weak and the findings across studies are inconsistent. The only socio‐demographic feature that is consistently linked with higher patient satisfaction with care is older age, which however is also associated with higher satisfaction with life in general.

More substantial correlations have been found with clinical characteristics and patient reported outcomes, such as subjective quality of life. Patients with higher symptom levels, especially more depressive symptoms, with personality disorders and with lower subjective quality of life tend to express less satisfaction with their care8.

Only a few aspects of care have consistently been found to impact on patient satisfaction. Coercive treatment and the perception of a negative therapeutic relationship are strongly associated with lower satisfaction with care, which might however be regarded as highly expected findings. There also seems to be a tendency for patients to be more satisfied with treatment in the community than in hospitals9.

When satisfaction scores are obtained for the evaluation of different treatments and services, one should consider the above determinants – e.g., age, the legal status of the treatment, and severity of illness or symptoms, in particular depressive symptoms – as potential confounders. Adjusting scores for these confounders minimizes the risk that positive or negative ratings get falsely attributed to a specific form of care when in fact they reflect general tendencies of a patient group with specific characteristics. For instance, patients with marked depressive mood are more likely to express lower satisfaction with any form of care.

Adjusting for age and the legal status of treatment should usually be feasible in mental health services, as such data are available in most routine data documentation systems. In many research studies, one can also obtain observer or self ratings of symptoms, including depressive symptoms. When patients rate their satisfaction in routine care, however, it is often not possible to assess their symptom levels at the same time. Still, considering some global rating of symptom severity would be helpful.

How patient satisfaction with mental health care should be assessed in research and practice depends mainly on the scope and purpose of the assessment. Quantitative scores as provided by the established scales can be helpful, if an adjustment for confounders is possible. Some scales are short and simple to use, and provide helpful scores for research studies and broader evaluations of services or treatments. When using the scales, one might, however, also want to be aware of their limitations.

When satisfaction scores are obtained to evaluate services, substantial differences of such scores between services or significant changes over time are unlikely, when all confounders are considered. Frequent measurement of satisfaction scores may, therefore, not be very informative. Also, differences on quantitative scores alone will not be a precise guide for which aspects of care should be improved to raise the satisfaction of patients. For this, one may want to analyze subscales or single items of scales. Even these scores, however, have limitations, as no scale covers all aspects of care, and low satisfaction scores do not necessarily indicate what exactly should be done to make patients more satisfied.

Better than quantitative scales, open questions on what specifically patients are satisfied or dissatisfied with can elicit information on a wide range of aspects of care that may be relevant in a given context and that professionals can potentially act on. For example, if patients express dissatisfaction with the behaviour of one particular staff member or with the timing of home visits or with the dose of their medication, clinicians may change these aspects of care and thus directly improve patient satisfaction.

Finally, no scale or survey can replace the most important procedure to assess patient satisfaction with care in practice, which is a direct and open communication between patients and clinicians about patients’ experiences, appraisals and wishes. This can facilitate ongoing consideration of these experiences and views in shared treatment planning and service development.