OUP user menu

A population-based survey using PPE-15: relationship of care aspects to patient satisfaction in Hong Kong

Eliza L.Y. Wong, Michael C.M. Leung, Annie W.L. Cheung, Carrie H.K. Yam, E.K. Yeoh, Sian Griffiths
DOI: http://dx.doi.org/10.1093/intqhc/mzr037 390-396 First published online: 29 June 2011


Objective Satisfaction with hospitalization may lead to both better health outcomes for patients and create better working environment and relationship between staff and patients. The objective of this study is to explore the relationship between the experiences of inpatients with specific aspects of care and satisfaction with the hospitalization experience.

Design A secondary data analysis based on the Thematic Household Survey which was regularly conducted by the Census and Statistic Department.

Setting This survey covered the land-based non-institutional population of the Hong Kong Special Administrative Region (HKSAR), China, and representing 99% of the total population of HKSAR of 6.8 million.

Participants Totally 1264 respondents aged ≥15 who had ever been admitted to local hospital in 12 months prior enumeration were recruited.

Main Outcome Measures The Picker Patient Experience Questionnaire-15 (PPE-15) was adopted for measuring patient-perceived quality of hospitalization; and one global rating of the overall quality of hospital service was included.

Results The mean global satisfaction scores for public and private hospital care were 7.3/10 and 7.8/10, respectively. By adjusting patient demographics, the regression models show that ‘want to be more involved in decision made about the care and treatment’, ‘respect for patient's dignity’, ‘patients' family have enough opportunity to talk to doctor’ and ‘tell about danger signals regarding illness/treatment after went home’ are major determinants of the global satisfaction scores.

Conclusion Communication, respect and patient engagement in provider–patient relationship are important in determining patient's satisfaction. Training and healthcare education curriculum could take this into account for ensuring the quality of patient-centered care.

  • patient satisfaction
  • quality measurement
  • quality indicators


There is growing evidence that patient satisfaction correlates with better quality of care [13]. Studies in the UK and USA have noted that hospitals with more satisfied patients generally provided higher quality of care as measured by validated quality metrics using standard methodology [2, 3]. Studies around the world have also noted that patient satisfaction is associated with increased compliance with the prescribed treatment and discharge instructions, reduction of complaints against the institution and improvement of morale and job satisfaction among health-care providers, which in turn can be of benefit to both patients' health outcomes and relationships with health-care professionals [48]. Patient's experience of care is becoming a more and more important indicator of quality of care [4, 9, 10]. Standard questionnaires such as Picker questionnaire [11], Hospital Consumer Assessment of Healthcare Providers and System [3, 12], Victoria Patient Satisfaction Monitor [13] have already been regularly incorporated in patient surveys in the UK, USA and Australia, respectively.

Several studies have shown that patient satisfaction can be associated with patient characteristics, including age [10, 1421], gender [15], race [10], education level [20] and health status [10, 16, 17, 22]. Global patient satisfaction is also influenced by quality dimensions such as type of health-care organization, tangible services and intangible services. A study in Turkey found that inpatients in the private hospitals were more satisfied with service quality than those in the public hospital [23]. Patients' participation in decision-making and the doctor–patient relationship were also found to be factors of importance; in general, a higher satisfaction was found in patients with perceived adequate sharing in decision-making and good doctor–patient relationships [24, 25]. Receiving information or clear explanation from health-care staff also contributed significantly to the overall patient satisfaction [15, 26]. However, most previous studies have only taken into account only one or two quality aspects and some have not adjusted for other patient characteristics.

Health-care system in Hong Kong

Similarly to the UK, there is a mixed hospital care system in Hong Kong, about 90% of in-patient services are managed by the tax-funded public sector. The Hospital Authority provides a comprehensive range of secondary and tertiary specialist care and medical and rehabilitation services to patients through 41 public hospitals, 48 specialist outpatient clinics, and 74 general outpatient clinics, which are organized into seven geographical clusters with a population catchment of ∼1 million people. In each cluster, there are one or more acute hospitals, which admit acutely ill patients from acute and emergency units, and one or more convalescence/rehabilitation hospitals which admit transfer from the acute hospitals for rehabilitation and convalescence. It is commonly believed by the general public that public hospital care is associated with lower patient satisfaction than the private hospital care due to the institutional characteristics such as limited manpower and long waiting times [23, 27]. A recent study measured self-reported inpatient experience in Hong Kong using the Picker Patient Experience Questionnaire-15 [28]. The finding confirmed results of a UK study: patients who sought care from private hospitals reported significantly higher satisfaction than those cared for in public hospitals, however, the underlying reasons for the different satisfaction between public and private hospitals and determinants of inpatient satisfaction were not explored. The determinants of satisfaction in the UK may differ from the determinants in other cultures. The goal of this study was to identify associations between global patient rating of hospital service quality and specific aspects of hospital care controlling for patient characteristics with the hope that this information can guide providers' improvement efforts.


Data collection

This study employed the Thematic Household Survey (THS) 2007 database which is a cross-sectional survey conducted regularly by the Census and Statistic Department, Government of the Hong Kong Special Administrative Region (HKSAR) of the People's Republic of China and it started in 1999. The survey enquires about statistical information on social topics proposed by individual government departments to form different rounds of THS and contracted-out to independent private research firms. Each round of survey is an independent and territory-wide survey covering the land-based non-institutional population of HKSAR, China and representing 99% of the total population of the HKSAR of 6.8 million. The round of THS 2007 covered six topics, including health status of Hong Kong residents, doctor consultation, hospitalization, dental consultation, provision of medical benefits by employers/companies and coverage of medical insurance purchased by individuals and knowledge, attitude and practice of medical checkups. In the survey, the respondents aged 15 and over who completed junior high school and lived independently (not needing parental/guardian consent) were approached for face-to-face interviews by trained enumerators using the structured questionnaire. Between November 2005 and March 2006, the survey included a total of 33 263 household individuals, constituting a response rate of 79.2%. The sub-sample THS household respondents aged ≥15 who had ever been admitted to local hospital in 12 months prior enumeration were recruited (n= 1264). This target sample was asked to assess their overall health-care experiences in their latest hospitalization experiences. We analyzed the most recently reported hospitalization episode to minimize the recall bias [29].

Survey assessment instrument

The previously validated short form Picker Patient Experience Questionnaire-15 (PPE-15) was adopted for measuring patient perceived quality of hospitalization [11]. The PPE-15 provides one global rating of the quality of health care and measures seven aspects of in-hospital patient experiences, including communication with doctors and nurses, anxiety or fear, decision engagement, respect and dignity, pain control, information available and danger sign for follow-up (Supplementary material, Appendix 1). For each item, three- and four-point scale responses were adopted. For the global rating of the health-care quality from the patient perspective/experience, the satisfaction score of 0–10 was applied, where the score 10 referred to the highest satisfaction.

Statistical analysis

Demographics of respondents and responses profile were descriptively presented. Bivariate analysis of demographics and the place of attendance and global satisfaction were performed. Regression analysis models were conducted with the global satisfaction rating as the dependent variable and independent variables including the specific aspects of hospital care, place of attendance (i.e. public or private hospital) and demographic and socio-economic characteristics. The first model included only place of attendance. Subsequent models included all aspects of hospital care together and the other independent variables listed above. All data management and analysis were performed using Stata Version 10.0.



In the 12 months prior to enumeration, 3.8% (n= 1264) of all respondents were admitted to the local hospital. For the latest episode of hospitalization, 73.9% (n= 934) consulted the public hospitals and 26.1% (n= 330) attended the private hospitals. Table 1 showed that the majority of the respondents were female (55.1%, n= 696), completed secondary school (43.2%, n= 546), married (64.5%, n= 815) and had not received government allowance (69.2%, n= 875); the average age was about 54.4 years old and 58.1 % (n= 734) of respondents had chronic illness. About 43% (n= 542) of the respondents admitted to hospital reported excellent/very good or good health status. By comparing the demographic profile of patients between public and private hospitals, patients admitted to public hospitals were older and with primary or lower education background (P< 0.001). Also, more patients were not married, had chronic illness, were receiving government allowance and rated their health as Fair or Poor (P< 0.001). Responding to the out-of-pocket hospital bill payment, patients in the public hospital paid UK$ 2969.8 (HK$23 164.6) less per year than those admitted in the private hospital (P< 0.001).

View this table:
Table 1

Demographic and health-related factors of respondents

Last episode of hospitalization in the past 12 months in Hong Kong and reported prior to enumeration, n= 1264Admitted to Public Hospital, n= 934, n (Col %)Admitted to private hospital, n= 330, n (Col %)Overall, n (Col %)P-valuea
Age, mean years (SD)57.6 (20.7)45.3 (15.1)54.4 (20.1)0.000*
 Male448 (48.0)120 (36.4)568 (44.9)0.000*
 Female486 (52.0)210 (63.6)696 (55.1)
 Primary or below446 (47.8)65 (19.7)511 (40.4)0.000*
 Secondary385 (41.2)161 (48.8)546 (43.2)
 Tertiary103 (11.0)104 (31.5)207 (16.4)
Marital status
 Not currently married360 (38.5)89 (27.0)449 (35.5)0.000*
 Currently married574 (61.5)241 (73.0)815 (64.5)
Chronic illness
 No330 (35.3)200 (60.6)530 (41.9)0.000*
 Yes604 (64.7)130 (39.4)734 (58.1)
Self-rated health status
 Excellent/very good52 (5.6)24 (7.3)76 (6.0)0.000*
 Good309 (33.1)157 (47.6)466 (36.9)
 Fair363 (38.9)112 (33.9)475 (37.6)
 Poor210 (22.5)37 (11.2)247 (19.5)
Received any government allowance
 No566 (60.6)309 (93.6)875 (69.2)0.000*
 Yes368 (39.4)21 (6.4)389 (30.8)
Out of pocket bill payment, mean (SD) [HK$]1118.9 (5278.9)24283.5 (17176.7)6098.6 (13255.8)0.000*
  • aChi-square tests/t-tests.

  • *Statistically significant at 0.05 level.

Patient satisfaction in public and private hospitals

The majority (77.9%) rated the global satisfaction with the hospital as score 7 or above, where 10 is the highest satisfaction score. The mean global satisfaction scores for public and private sectors were 7.3 and 7.8, respectively. The result indicates that there were significant differences in public and private sectors (P< 0.001). Table 2 shows that 38% of the patients admitted to public hospitals reported excellent/very good or good health status compared with 55% among the patients admitted to private hospitals (P< 0.001). Patients who paid more out of pocket payment (P< 0.001) did not have chronic illness (P= 0.009) and reported higher health status (P= 0.019) were more satisfied with the hospital services.

View this table:
Table 2

Relationship between global satisfactiona and patient demographic

Patient demographicsMean of global patient satisfactionP-value
Admitted to
 Public Hospital7.260.000*
 Private Hospital7.84
 70 and above7.40
Out of pocket payment (HK$)
 1000-under10 0007.42
 ≥10 000 and above7.82
 Primary or below7.360.430
Marital status
 Not currently married7.310.071
 Currently married7.47
Chronic illness
Self-rated health status
 Very good7.76
Received any government allowance
  • aPatient satisfaction is a global rating on the quality of health services, where the maximum score is 10.

  • *Statistically significant at 0.05 level.

Determinants of patient satisfaction

In the multivariable model including all characteristics reported in Table 2, the coefficients for the place of hospitalization, whether in the public and private hospitals, were not significant. Table 3 shows that four aspects of care including patient involvement (P= 0.005 for some extent; P= 0.007 for definitely), respect for patient dignity (P= 0.006 for some extent; P< 0.001 for definitely), availability of doctor to talk to carers (P= 0.001 for definitely) and explanation of dangerous signs after discharge (P= 0.004) were all associated with the global patient satisfaction score. Respect for patient dignity had the largest coefficient (R= 0.295–0.219).

View this table:
Table 3

Multivariate linear regression of association between global patient satisfactiona and nature of health-care organization

Independent variablebModel 1cModel 2c,d
(R2 = 0.028)(R2 = 0.207)
Beta (SE)95% (CI)P-valueBeta (SE)95% (CI)P-value
Admitted to private hospital0.102 (0.050)0.003, 0.2010.043*0.010 (0.047)−0.082, 0.1030.825
Want to be more involved in decision made about the care and treatment
 Yes, definitely−0.121 (0.045)−0.211, −0.0330.007*
 Yes, to some extent−0.125 (0.044)−0.211, −0.0380.005*
Feel that were treated with respect and dignity while in hospital
 Yes, definitely0.295 (0.082)0.133, 0.4560.000*
 Yes, to some extent0.219 (0.080)0.062, 0.3750.006*
Family or someone else close have enough opportunity to talk to a doctor
 Yes, definitely0.160 (0.050)0.062, 0.2580.001*
Staff member of the hospital tell about danger signals regarding illness or treatment to watch for after went home
 Yes, definitely0.100 (0.047)0.008, 0.1910.034*
  • Beta, unstandardized coefficient; SE, standard error; CI, confidence interval.

  • aGlobal patient satisfaction was transformed and the transformed scale is 8.64–11 where higher score means higher satisfaction.

  • bBesides the variable ‘Admitted to private hospital’, only statistically significant variables were reported in the above table.

  • cAll models were adjusted by age, gender, educational level, marital status, whether had a chronic illness, self-perceived heath, whether received government support and the bill payment for the admission which were listed in Table 2.

  • dModel 2 reported the regression result of association between global patient satisfaction and all variables derived from the 15 items of the Picker questionnaire.

  • *Statistically significant at 0.05 level.


This is the first study to explore the relationship between specific aspects of hospital care and patients' global rating of hospital satisfaction in Hong Kong. The results indicate the utility of these measures for monitoring the performance of hospitals in the future. In Hong Kong, the community-dwelling population is generally satisfied with hospital care, having a rating score of 7 or above where 10 was the highest, however, these scores are a bit lower than those noted in the USA where the majority of patients rate hospital care as 9 or 10 of 10 [3]. In the UK, the majority of inpatients rated the hospital services as good to very good on a 5-point scale ranging from poor to excellent, but comparison with the Hong Kong results is difficult because of different scale [30].

Comparison of public and private health service sectors using patient satisfaction data is becoming common. This study shows that in Hong Kong the level of patient satisfaction with hospital care obtained from the public and private hospitals was not significantly different after statistical adjustment. In contrast to studies from Turkey and Bangladesh [23, 27], the satisfaction ratings for the public hospitals were significant lower than private hospitals.

The quality and outcomes framework has a focus of targeting incentives on more holistic aspects of care, including patient experiences [31]. In addition to clinical outcomes and tangible care, interpersonal care was found to be an important correlate of patient satisfaction rating including patient involvement in the care and treatment, effective answers by doctors, respect for patient dignity and information on dangerous signs after discharge. These factors were associated with patient satisfaction independent of the setting in which health services were received.

The findings suggest that the provider–patient relationship had major influences on patient satisfaction, which agreed with findings in the USA and Singapore that a strong doctor–patient relationship and adequate communication between the health-care professionals and patients were important factors for patient satisfaction [12, 32, 33]. Another study has shown that perceived a good communication from the patient's perspective was simple, easily understood information presented in a timely fashion from doctors or nurses [6]. Adherence to treatment and better health outcomes are both associated with better provider–patient relationships that include sufficient explanation of the care-related information [24, 32]. Patient waiting in reception areas without adequate information correlated negatively with patient satisfaction in a recent UK study [34]. Our results suggest that patients value involvement in decision-making about care and treatment. This shared decision-making will require improvements in communication skills among health-care provider.

In our study, respect for the patient's dignity was an important determinant of satisfaction. Other studies from across the globe have found that patient values, preferences and dignity were found to be significantly related to the reported satisfaction in both in-patient and out-patient health settings [3, 31, 35]. Our finding signals that patient involvement during the hospitalization was not sufficient. This is consistent with other western studies which tend to find that patients want to be involved in sharing decision-making about their treatment plan [25, 34, 36]. On the one hand, patients may be confident that they can communicate with their providers about their views of treatment, while on the other hand, patients may not feel they have enough involvement during hospitalization. If health-care workers' explanations were easier to understand, patients might feel more empowered to engage in treatment decisions [37]. A good medical decision should engage patient with absence of pressure during the decision process [37]. Our results also suggest that patients value relevant information about the potential danger signs to be aware of after discharge. Poor discharge instructions may exacerbate hospital readmissions [3].

This study has some limitations. The survey lacked questions about waiting times and nurse/doctor-staffing levels which are known to be important factors contributing to the reported satisfaction [3, 38, 39]. Insurance coverage was not included in the regression models because the sample size was small and this variable was highly correlated with place of attendance. Participants were asked about hospitalization experiences during the past 12 months, so recall bias could affect our results. Jackson and Verberg [41] suggested that with a longer time lag between care and survey response, patient satisfaction rises so a recall bias could inflate global satisfaction scores in our study [40]. Self-interest bias might also exist if participants perceive that expressions of satisfaction will enhance their future receipt of services. Information on the reason of admission, and severity or co-morbidity was not captured and these factors may be associated with the patient satisfaction.

In light of the growing number of efforts in the UK, USA and Australia to assess the service quality of health-care organization on the basis of patient satisfaction data, Hong Kong has moved to implement a routine national standard patient satisfaction survey. These surveys are intended to guide professional efforts to improve the quality of health care, allow comparison between different clusters within Hong Kong and international settings and engage the patients in the quality improvement process. The findings can serve as feedback to the hospital staff on the care provided and encourage them to improve patient satisfaction further [12]. With the growing importance of patient satisfaction as an indicator of quality of care [4, 9, 10], it is important to identify the aspects of care that are essential to patients by controlling for patient characteristics.


Our results indicate that patients are generally satisfied with hospital care. Some aspects of care are highly correlated with patients' reported satisfaction, including patients' involvement in their care and treatment, availability of doctors for patients' families, respect for patient dignity and explanation of dangerous post-discharge signs. As one of the first studies of this topic in the hospital setting in Hong Kong, these findings can enrich the implementation of a national standard for patient satisfaction in health care in Hong Kong and serve as a benchmark for other countries with similar health systems.


We would like to thank the Census and Statistic Department of the Government of the HKSAR, which provided us with the data for this study.


View Abstract