Does Patient Satisfaction Reflect Differences in Care Received After Stroke?
Background and Purpose—Despite the volume of research into patient satisfaction, it is not clear whether satisfaction ratings reflect differences in care received after stroke or the characteristics of patients. The aim of this study is to test the hypothesis that patient satisfaction is independently related to differences in care received after stroke.
Methods—Stroke patients participating in a randomized controlled trial of early discharge to community therapy completed a satisfaction questionnaire and physical and psychological outcome measures at 4 and 12 months. Two hundred seventy-four patients (83%) were followed up at 4 months and 262 patients (79%) at 12 months. Use of therapy and community services was quantified. Logistic regression was used to investigate associations between satisfaction, patient characteristics, and service provision.
Results—The more therapy, meals on wheels, and home help visits patients received, the more likely they were to be satisfied. Patients in the conventional arm of the trial were less likely to express satisfaction than those discharged early to community therapy. Women, older people, anxious people, and those more functionally independent were more likely to be satisfied, while those with depression, speech and swallowing deficit, motor deficit, and poor subjective health were less likely to be satisfied.
Conclusions—The finding that satisfaction assessments reflect real differences in the provision of care and occur independently of any associations with patient characteristics is new. Patient satisfaction should be taken seriously as an outcome.
Research into patient satisfaction after stroke spans the last 5 years only. Six1 2 3 4 5 6 of the 9 studies we found1 2 3 4 5 6 7 8 9 used the same questionnaire, which examines satisfaction with 12 aspects of stroke care.1 This questionnaire has been examined for test-retest repeatability, internal consistency, convergent and discriminant validity, content, and construct validity and has been found to be valid and reliable. Questions can be analyzed individually or summed to produce a total score.
When this questionnaire was used, the highest rates of dissatisfaction 6 months after stroke were with the amount of therapy, the amount of recovery made, information received, and services after discharge.1 2 These studies, however, did not relate satisfaction ratings to the actual services patients received. Consequently, it was not known whether, for example, dissatisfaction with therapy was related to having received less therapy. Studies that used the questionnaire to investigate the outcome of differences in service provision have been more informative. When satisfaction was compared between stroke patients who had received family support and those who had not, those with support were more satisfied with the information received.4 Similarly, a randomized controlled trial of a stroke family care worker,5 which used an extended version of the questionnaire, found the intervention group to be significantly more satisfied with the care received. Recently, a randomized controlled trial of early discharge to community therapy6 found no differences between the 2 groups in terms of physical or psychosocial outcomes, but when the questionnaire was used, patients in the intervention group were significantly more satisfied with aspects of their care.
While these findings seem to indicate that satisfaction is related to differences in care provision, the influence that patients’ characteristics could have on their expressions of satisfaction is unclear. For example, a consistent finding in the wider satisfaction literature is that older people are more likely than younger people to express satisfaction with care.10 11 In the field of stroke, dissatisfaction has been associated with depression1 and “emotional distress,”8 although findings on the relationship between satisfaction and functional ability have been conflicting.1 8 Only one study has attempted to investigate the relationship between patient satisfaction, stroke services received, and patient characteristics, but this study used an aggregate satisfaction score and was therefore unable to identify whether, for example, satisfaction with general practitioner care was related to specific aspects of general practitioner care.8 Consequently, it is unclear whether patient satisfaction after stroke is independently related to services received. In the general field of patient satisfaction, the relationship between satisfaction, patient characteristics, and care received is also unclear.12 13 14 15 This study explores whether patient satisfaction reflects differences in care received or patient characteristics. The hypothesis is that patient satisfaction is independently related to differences in care received after stroke.
Subjects and Methods
Data from the randomized controlled trial of early discharge to community therapy after stroke are used to address this question. The methodology of the trial has been described elsewhere6 and will be presented only briefly here. Stroke patients were recruited from hospital stroke registers between January 1993 and July 1995. They were managed as usual until they were able to either transfer independently (if living alone) or with assistance (if living with a caregiver). They were then randomized to either continue conventional treatment or to be discharged early and receive treatment in the community.
Patients in the conventional care arm of the trial received therapy as usual. Those in the early discharge group were given a planned course of domiciliary physiotherapy, occupational therapy, and speech therapy, with visits as frequently as considered appropriate (the maximum being 1 daily visit from each therapist). Patients received care from the team for a maximum of 3 months. The community team comprised a senior physiotherapist, a senior occupational therapist, a half-time speech therapist, and a full-time therapy aide. Outpatient resources available to patients in both arms of the trial included a stroke clinic, generic domiciliary physiotherapy and speech therapy, outpatient physiotherapy, and the usual community resources.
The number of units of therapy (all types) received by patients was categorized as 0 to 14 units (reference category), 15 to 28 units, 29 to 63 units, or >63 units (1 unit of therapy=20 minutes). These categories were chosen to approximately represent 1 unit per day for <2 weeks, 2 to 4 weeks, 1 to 2 months, or >2 months. Contact with meals on wheels, home help, district nurse, lunch club, or hospital physician was recorded.
Patient Characteristics: Demography and Outcomes
Baseline data on demography and stroke severity were collected within 48 hours of randomization. We assessed outcomes at discharge and 2, 4, 6, and 12 months after stroke using a range of standardized assessments16 : the Mini-Mental State Examination (MMSE) has a maximum score of 30, with a higher score indicating less cognitive impairment; the Motricity Index has a maximum score of 100, with a higher score indicating greater function; the Barthel Index has a maximum score of 20 and the Rivermead Activities of Daily Living Score is scored from 15 to 45, with less disabled people scoring higher on both; the Hospital Anxiety and Depression Scale has a maximum score of 21, with more anxious and depressed people scoring higher; the Nottingham Health Profile (NHP) has a maximum score of 45, with a higher score indicating poorer subjective health. The Barthel scale was divided into 0 to 14, 15 to 19, and 20 to overcome differences between the intervals at the lower, middle, and upper ends of the scale. All other scales were analyzed as continuous variables.
We assessed patient satisfaction by interview using the questionnaire developed by Pound et al1 at 4 and 12 months only. Patients were asked to agree, strongly agree, disagree, or strongly disagree with each of 12 statements about their care, which can be grouped into the areas of inpatient care, therapy/recovery, and services after discharge (see Appendix).
The questionnaire responses were collapsed into 2 categories of “agree” and “disagree.” Rather than using a summary score, analysis was by individual question to allow a detailed examination of the relationship between satisfaction and care received.
Because satisfaction was assessed at 4 and 12 months only, these are the only time points at which the other outcomes are considered for this analysis. We used χ2 tests (categorical outcomes) and unpaired t tests or Mann-Whitney tests as appropriate (continuous outcomes) to investigate the univariate relationships between each satisfaction question and both demographic variables and outcomes at each time point. The McNemar test was used to examine differences between satisfaction at 4 and 12 months.
A multivariable logistic regression model with backward stepwise elimination (P=0.1) was used to obtain a final model for each satisfaction question at both 4 months and 1 year. Variables considered for all satisfaction questions were the Barthel Index, NHP, Hospital Anxiety and Depression Scale, MMSE, Motricity Index, Rivermead Activities of Daily Living, arm of trial (conventional care or early discharge), age, and sex. For satisfaction questions that were related to specific services used, variables on the relevant service were added to the model (categorized as described above). Deficit (presence of weakness or paralysis in face or any limb) and speech/swallowing deficit at the appropriate time point were also included for the questions on therapy and recovery.
Interactions between explanatory variables were not examined because of the large number of comparisons being made.
Two hundred seventy-four patients of a possible 331 (83%) were followed up at 4 months, of whom 201 (73%) completed a satisfaction questionnaire (42 died, 2 moved, 7 were lost to follow-up, and 6 refused). Two hundred sixty-two patients (79%) were followed up at 1 year, of whom 194 (74%) completed a satisfaction questionnaire (60 died, and 9 were lost to follow-up).
Overall Levels of Satisfaction
At 4 months after stroke, there were high rates of satisfaction with inpatient care, lower satisfaction with the amount of information received and amount of recovery made, and very low satisfaction with the amount of therapy received. The same pattern was found at 12 months, with no significant changes over time (Table 1⇓).
Patient Characteristics: Outcomes and Demography
As described elsewhere,6 there were no significant differences between patients in the 2 arms of the trial at discharge. (Tables showing demographic variables, outcomes, and their relationship to satisfaction are available upon request.)
The mean number of days in the hospital from randomization was 12 for the intervention group and 18 for those receiving conventional care (P=<0.01). Satisfaction with services was univariably related to use of some services at both 4 and 12 months (Table 2⇓).
Multivariable Analysis: Relationship Between Satisfaction, Patient Characteristics, and Services Received
Satisfaction with inpatient care was related to care received. At 12 months, patients in the conventional care group were less likely than patients in the early discharge group to be satisfied with most aspects of inpatient care (Table 3⇓). Satisfaction with inpatient care was also related to patient characteristics, with depressed patients being less likely and anxious patients being more likely to express satisfaction with inpatient care at 4 months. At 1 year, patients with lower subjective health (according to the NHP) were less likely to express satisfaction, while more independent patients (according to the Rivermead Activities of Daily Living score) were more likely to express satisfaction with aspects of inpatient care.
Therapy and Recovery
Satisfaction with therapy was related to the amount of therapy received; patients receiving from 15 to 28 units of therapy were more likely to be satisfied at 4 months than those receiving ≤14 units, while those receiving >63 units were less likely to be satisfied than those receiving ≤14 units (Table 4⇓). Patients in the conventional care arm of the trial were less likely to be satisfied with the amount of recovery made at 1 year than those in the early discharge group.
Satisfaction with therapy and recovery was also related to patient characteristics; at 4 months patients with any kind of deficit and at both time points patients with poorer subjective health were less likely to be satisfied with the amount of therapy received. At both time points women were more likely to be satisfied with the amount of therapy received, while at 1 year patients with a higher Motricity Index score, those who were more anxious, and those with a higher Rivermead score were more likely to be satisfied with the amount of therapy received. Patients with any kind of deficit and those who were depressed were less likely to be satisfied with the amount of recovery at both time points, while those with a higher Rivermead score were more likely to be satisfied with the amount of recovery made at 1 year.
Services After Discharge
Satisfaction with services after discharge was associated with the actual services received (Table 5⇓). Patients who received meals on wheels (4 months) and visits from home help (12 months) were more likely to be satisfied with the information they received about services after discharge and with the support received at home. At 1 year patients in the conventional care arm of the trial were less likely to be satisfied with the amount of contact they had with the hospital after discharge. Depressed patients and those with lower subjective health were less likely to be satisfied with services after discharge, while women, anxious patients, and those with a higher MMSE score were more likely to be express satisfaction with most aspects of services after discharge. The older patients were, the more likely they were to be satisfied with the information they received about services in the community at both time points.
The response rate for the satisfaction questionnaire was as expected for a questionnaire requiring subjective responses from a population with high levels of cognitive impairment.1
The trial was not designed with patient satisfaction as a primary outcome; rather, the sample size was based on detecting a difference in Barthel Index at 12 months between the 2 treatment groups. The achieved sample sizes for the satisfaction questionnaire (201 at 4 months and 194 at 12 months) limit the study’s power to detect differences in satisfaction between the 2 arms of the trial. With a power of 80% and a significance level of 5%, the detectable proportion satisfied in the conventional group is between 74% (where satisfaction in the early discharge group is 90%) and 30% (where satisfaction in the early discharge group is 50%). Because the level of satisfaction is high (>80% for 7 of the 4-month and 5 of the 12-month questions), the ability to detect associations between covariates and satisfaction is low. For the same reason, the predictive power of any associations found will also be low. The large number of associations being tested means that some will appear to be significant by chance. Despite these drawbacks, this study provides an indication of the variables associated with patient satisfaction in a reasonably large cohort of stroke patients.
The finding that women and older people were more likely to express satisfaction with some aspects of their care is not confined to the field of stroke.10 11 Women’s greater likelihood of being satisfied with services in the community may reflect their low expectations in this area, since they have traditionally filled gaps in community services. The associations between depression, poor subjective health, deficit, and decreased satisfaction, on the one hand, and between functional independence and increased satisfaction, on the other, support findings outside the field of stroke that patients with better health status are more satisfied.12 Anxious patients’ increased likelihood of expressing satisfaction may reflect a desire to please among these patients or a fear of being seen as “complaining.”
The most important finding is that patients’ assessments of their care reflected real differences in the provision of care and occurred independently of any associations with demographic or outcome variables. The association between increased satisfaction and increased support from community services makes sense intuitively, but the relationship between satisfaction and the amount of therapy is more complex. People with stroke often give the impression that they never want therapy to stop,17 yet while an increase in therapy was positively associated with satisfaction among the patients in this sample, they also indicated a ceiling for the amount of therapy they appreciated. Further, qualitative research is necessary to investigate why this might be.
Patients were not only concerned with quantity, however, but also with the way in which services were delivered. Patients in the intervention group were more satisfied with aspects of care in which the intervention would be expected to have an effect, such as information received, contact with the hospital after discharge, the amount of recovery made, and satisfaction that physicians had done everything possible. Possibly the conventional care patients’ greater dissatisfaction with the way their “personal needs” were met in the hospital was due to their longer admission and the greater opportunity they had for “dissatisfying events” to occur.
The results suggest that the questionnaire is sensitive to changes in the provision of care, but the wider implication is that patients’ assessments should be taken more seriously than at present. As noted above, the same patient satisfaction questionnaire1 was recently adapted and used in a randomized controlled trial of the benefits of a support worker after stroke.5 Although patients in the intervention group were significantly more satisfied in areas in which the support worker would be expected to have an effect, the service was withdrawn. Presumably because patients’ assessments were not in accordance with the “objective” (ie, professionally defined) outcomes, the patients’ assessments were accorded little weight. We suggest that the results from our randomized controlled trial provide support for early discharge to community rehabilitation, despite the lack of significant differences between the 2 groups in terms of impairment and disability.
In conclusion, this study, in providing evidence that patient satisfaction reflects real differences in the provision of care and is independent of patient characteristics, supports the plea that patient satisfaction should be taken more seriously as an outcome measure after stroke.
Patients are asked to agree, strongly agree, disagree, or strongly disagree with each of the following statements:
1. I have been treated with kindness and respect by the staff at the hospital.
2. The staff attended well to my personal needs while I was in the hospital (eg, I was able to get to the toilet whenever I needed).
3. I felt able to talk to the staff about any problems I might have had.
4. I have been given all the information I want about the causes and nature of my illness.
5. The doctors have done everything they can to make me well again.
Therapy and Recovery
6. I am satisfied with the type of treatment the therapists have given me (eg, physiotherapy, speech therapy, or occupational therapy).
7. I have had enough therapy (eg, physiotherapy, speech therapy, or occupational therapy).
8. I am happy with the amount of recovery I have made since my illness.
Services After Discharge
9. I was given all the information I wanted about the allowances (eg, welfare benefits) or services (eg, home help, meals on wheels, district nurse) I needed after leaving the hospital.
10. Things were well prepared for my return home.
11. I get all the support I need from services such as meals on wheels, home help, and district nurses.
12. I am satisfied with the amount of contact I have had with the hospital since I have been discharged.
This study was supported by the Stroke Association; Lambeth, Southwark, and Lewisham Health Authority; Special Trustees of St Thomas’s Hospital; Nuffield Provincial Hospitals Trust; and Wandsworth Health Gain Fund.
- Received May 26, 1998.
- Revision received October 2, 1998.
- Accepted October 2, 1998.
- Copyright © 1999 by American Heart Association
Pound P, Gompertz P, Ebrahim S. Patients’ satisfaction with stroke services. Clin Rehabil.. 1994;8:7–17.
Gompertz P, Pound P, Briffa J, Ebrahim S. How useful are non-random comparisons of outcomes and quality of care in purchasing hospital stroke services? Age Ageing. 1995;24:137–141.
Dennis M, O’Rourke S, Slattery J, Staniforth T, Warlow C. Evaluation of a stroke family care worker: results of a randomised controlled trial. BMJ. 1997;314:1071–1077.
Rudd AG, Wolfe CDA, Tilling K, Beech R. Randomised controlled trial to evaluate early discharge scheme for patients with stroke. BMJ. 1997;315:1039–1044.
Wellwood I, Dennis M, Warlow C. Patients’ and carers’ satisfaction with acute stroke management. Age Ageing. 1995;24:519–524.
Scholte op Reimer WJM, de Haan RJ, Limburg M, van den Bos GAM. Patients’ satisfaction with care after stroke: relation with characteristics of patients and care. Qual Health Care. 1996;5:144–150.
O’Mahony PG, Rodgers H, Thomson RG, Dobson R, James OFW. Satisfaction with information and advice received by stroke patients. Clin Rehabil. 1997;11:68–72.
Hall JA, Dornan MC. Patient sociodemographic characteristics as predictors of satisfaction with medical care: a meta-analysis. Soc Sci Med. 1990;30:811–818.
Ross Davies A, Ware JE. Involving consumers in quality of care assessment. Health Aff (Millwood). 1988;7:33–48.
Hall JA, Dornan MC. Meta-analysis of satisfaction with medical care: description of research domain and analysis of overall satisfaction levels. Soc Sci Med. 1988;27:637–644.
Stojcevic N, Wilkinson P, Wolfe C. Outcome measurement in stroke patients. In: Wolfe C, Rudd A, Beech R, eds. Stroke Services and Research. London, England: The Stroke Association; 1996.
Pound P, Bury M, Gompertz P, Ebrahim S. Views of survivors of stroke on benefits of physiotherapy. Qual Health Care. 1994;3:69–74.