Similarities among meaures of patient satisfaction
This study indicates that some hospital characteristics do not change the similarities of domains among patient satisfaction even though the same characteristics influence individual score differences within traditional analyses. In the present study, the nine HCAHPS measures were divided into three groups plus cleanliness, one group including doctor and nurse communications; a second group included pain management and two global evaluation measures; and a third group included medication explanation, quietness and staff responsiveness. Our study showed that hospital characteristics did not alter the similarities in these group but did affect the combinations of these groups. This study also suggested dissimilarities among the measures with low scores (medication explanation, quietness and staff responsiveness) and dissimilarities between overall rating and hospital recommendation.
Doctor communication and nurse communication were similarly assessed at all hospitals, with more “High” scores than all other measures. Pain management and the two global evaluation measures received similar scores but were dependent on hospital characteristics. These three domains, together with doctor and nurse communication, formed the higher-rated group at hospitals having larger SRSs as well as in those on the cardiac/general surgery and nursing/stroke care registries. At the other hospitals, however, pain management and the two global evaluation measures joined medication explanation, quietness and staff responsiveness to form the lower-rated group.
Studies on communication with medical staff have yielded contradictory results, with some reporting better communication with nurse practitioners than with doctors [34], and others reporting better communication with doctors than with nurses [6,12], and better communication with staff members at smaller than at larger hospitals [6,12]. Our study, however, found that scores on communication with doctors and nurses were similar and relatively higher than other measures at all hospitals, independent of hospital characteristics. Patients likely appreciate human contact, regarding communication with health care personnel as a sign of respect and a tool to meet their care needs and for avoiding possible medical malpractice [35].
Although overall rating shows a stronger correlation with nurse communication than with other measures including pain management [6,9], our study found that the similarities between overall rating and nurse communication were not consistent, but were limited at large hospitals providing acute surgical treatment. This suggests that the value of communication of medical staff to patients differ according to a patient’s medical status. Patients in pain view communication as a verbal and attitudinal aspect of care, with scores similar to those of global evaluation, whereas patients in less pain not requiring specialized treatment were highly appreciative of communication with doctors and nurses, but rated other hospital services as poor. This is likely an example of direct association among caring attitudes, swift pain treatment and patient satisfaction [17,37].
Previous studies have reported that scores on medication explanation, quietness and staff responsiveness differ according to hospital characteristics [7,13,36]. However, our study found that these three measures received similar poor ratings at all hospitals. The backgrounds of these low evaluations differed, as correspondence analysis showed that medication explanation had a “Low” rating, quietness had a “Medium” rating and staff responsiveness had an intermediate rating. These results were supported by MDS, suggesting their dissimilarities. A qualitative study reported that differences in scores on staff responsiveness and quietness may be due to differences in patient expectations, as patients are more tolerant of slow responses than of hospital noise, as they seek a quiet environment [35]. Although medication explanation and communication measures would seem to be related, asking in the patient HCAHPS questionnaire if explanations were easy to understand [30], no similarities were shown. Our study indicates that providers’ efforts to explain medications to patients were insufficient, possibly due to patient worries about the possibility of serious side effects including death [35], and the doctors failing to fully describe a medication’s side effects [36].
Overall rating and hospital recommendation have been treated equally in patient-satisfaction studies. Although they show a strong correlation in HCAHPS studies [6], our study suggests their possible dissimilarities, as MDS analysis placed them at some distance from one another, indicating patients regarded overall rating and hospital recommendation as being distinct, requiring further investigation.
Cleanliness was one measure that differed among groups of hospitals. For example, ACHs and CAHs showed similar results, except for cleanliness, as did hospitals outside the cardiac/general surgery registries and those outside the nursing/stroke care registries. ACHs may be a mixture of the three categories of SRSs, as ACHs represent between one-third and one-half of hospitals with smaller SRSs. Hospitals outside the nursing/stroke care registries may be less able to control the quality of medical care or have fewer resources as they do not submit process and outcome data to the government. However, these results are not due solely to the attitude toward quality control, as cleanliness is not the only measure poorly rated at these hospitals [8,11]. A qualitative study reported that patients regard lack of cleanliness as a possible indicator of infection [34] and bivariate analysis has shown relationships between cleanliness and technical quality [8,11]. The reasons for differences in cleanliness among groups of hospitals require further investigation. While previous studies have reported mixed results on whether EHR usage influences patient satisfaction [14,38], our study found that it did not.
Candidate factors for the structure of patient satisfaction
The results of this study suggest that hospital size, hospital type, the ability to provide acute surgical treatment and hospital interest in improving the quality of medical care were factors that may influence the structure of patient satisfaction, whereas EHR usage was not.
It should be noted that our analyses utilized adjusted percentages of hospital-level data. Analyses of real numbers are easier to grasp but have the disadvantage of being more heavily influenced by larger numbers. In our study, over 75% of all hospitals had SRSs ≥300. Using percentages can avoid the disadvantages of data imbalances. However, whereas the questionnaire developed from the HCAHPS survey data compared differences between hospital-level and individual-level findings [39], our methods did not. Individual-level data may produce different results.
Improving patient satisfaction
To improve patient satisfaction more attention should be paid on relationships among hospital services and their backgrounds in order to understand the depth of patient satisfaction. Hospitals should focus on medication explanation, noise reduction and rapid staff response, especially at large hospitals providing acute surgical treatment. At smaller hospitals, the improvement of pain management may lead to improvements in overall rating. Investigating the backgrounds of these groups will enhance understanding of patients’ viewpoints and behavior, thereby improving the quality of medical care.