International Journal for Quality in Health Care 1999; Volume 11, Number 4: pp. 309–317
Patient satisfaction in Bangkok: the impact of hospital ownership and patient payment status VIROJ TANGCHAROENSATHIEN1, SARA BENNETT2, SUKALAYA KHONGSWATT3, ANUWAT SUPACUTIKUL1 AND ANNE MILLS2 1
Health Systems Research Institute, Bangkok, Thailand, 2Health Economics and Financing Programme, London School of Hygiene and Tropical Medicine, London, UK and 3Ministry of Public Health, Bangkok, Thailand
Abstract Introduction. Patient satisfaction with care received is an important dimension of evaluation that is examined only rarely in developing countries. Evidence about how satisfaction differs according to type of provider or patient payment status is extremely limited. Objective. To (i) compare patient perceptions of quality of inpatient and outpatient care in hospitals of different ownership and (ii) explore how patient payment status affected patient perception of quality. Methods. Inpatient and outpatient satisfaction surveys were implemented in nine purposively selected hospitals: three public, three private for-profit and three private non-profit. Results. Clear and significant differences emerged in patient satisfaction between groups of hospitals with different ownership. Non-profit hospitals were most highly rated for both inpatient and outpatient care. For inpatient care public hospitals had higher levels of satisfaction amongst clientele than private for-profit hospitals. For example 76% of inpatients at public hospitals said they would recommend the facility to others compared with 59% of inpatients at private for-profit hospitals. This pattern was reversed for outpatient care, where public hospitals received lower ratings than private for-profit ones. Patients under the Social Security Scheme, who are paid for on a capitation basis, consistently gave lower ratings to certain aspects of outpatient care than other patients. For inpatient care, patterns by payment status were inconsistent and insignificant. Conclusions. The survey confirms, to some extent, the stereotypes about quality of care in hospitals of different ownership. The results on payment status are intriguing but warrant further research. Keywords: patient satisfaction, payment status, private sector health care
The Thai health care system is typical of many East and South East Asian countries in having a mixed economy of health care. Until recently, policy makers had focused their efforts on the public sector. However rising health care costs, increasing use of high technology equipment, and complaints of unethical practices at private hospitals have drawn policy makers’ attention to the mixed market for health care services in Thailand [1,2]. An emerging literature on quality of care in the private sector in developing countries has addressed mainly clinical quality of care at the primary level [3]. In terms of consumer perception of quality of care in the private sector, very limited evidence is available, but popular
stereotypes suggest that private (for-profit) providers are more likely to respond to patient preferences (particularly in terms of augmenting hotel aspects of care), whilst public sector providers may be less sensitive to patient preferences but more concerned with clinical quality [4]. More economically developed countries in East Asia, such as Japan, Korea and Taiwan, have attempted to bind public and private providers into a unified health care system through the extension of a Bismarckian model social insurance system (N. Ikegami, T. Hasegawa, unpublished manuscript 1990 [5–7]); Thailand plans to follow a similar path. There are currently several parallel insurance schemes in Thailand including the
Address correspondence to Sara Bennett, Health Economics and Financing Programme, Health Policy Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK. Tel: +44 171 927 2176. Fax: +44 171 637 5391. E-mail:
[email protected]
1999 International Society for Quality in Health Care and Oxford University Press
309
V. Tangcharoensathien et al.
Social Security Scheme (SSS) which provides health and other benefits to those in private, formal sector employment, the Civil Servants’ Medical Benefit Scheme (CSMBS) covering the medical expenditure for government employees, the Health Card scheme (a voluntary health insurance scheme) and the Low Income Scheme which provides fully subsidized access to public health care services for the poor. The issue of consumer satisfaction with care received is particularly pertinent in Thailand because of the manner in which health care providers under the SSS are paid. The hospital with which insured persons register (known as the ‘main contractor’, which may be publicly or privately owned) is paid a capitation fee: i.e. a lump sum annual amount for every insured worker choosing to register with that facility. In return for this amount the hospital must provide to all registered beneficiaries, all services in a defined package which excludes only a handful of extremely high cost interventions. Other medical benefit schemes (such as the CSMBS and private health insurers) pay providers on a fee-for-service basis. Both public and private hospitals charge fees to uninsured patients which are paid out-of-pocket. Beneficiaries of the SSS can change their main contractor hospital on an annual basis, so in principle there should be competition between facilities to attract more registered patients. However there are concerns that this competitive mechanism may not be sufficient to compensate for financial incentives to reduce the cost of care for insured persons. The SSS specifies certain structural standards which a facility must meet in order to be able to register as a main contractor under the scheme; however these structural standards do little to ensure the quality of process aspects of care. The incentives within the payment mechanism to lower the cost of care may be passed on to patients in the form of fewer investigations, fewer admissions and shorter consultation times. Attending physicians may come to see SSS patients as ‘less valuable’, and therefore worthy of less attention, than regular patients. Consumer satisfaction with quality of care may differ in some respects from an objective appraisal of quality; however, monitoring consumer satisfaction with quality of care is an important complement to other efforts to monitor quality in Thailand. This paper describes a study evaluating both inpatient and outpatient satisfaction with quality of care in nine hospitals of different ownership in Bangkok. The results reported are part of a larger study which examined differences in quality of care and efficiency in the nine hospitals and how method of paying for care affected these variables. With respect to consumer satisfaction the study aimed to: (i) identify differences in consumer satisfaction related to quality of care at the nine different hospitals; we were interested in how consumer perceptions corresponded to popular stereotypes of differences in care provided by hospitals of different ownership; (ii) examine differences in consumer satisfaction with quality of care by consumer payment status: our hypothesis was that SSS patients paid on a capitation basis may experience less courteous treatment, and hence be less
310
satisfied than those patients paying on a fee-for-service basis.
Development of survey instrument While there is a growing body of literature on the use of consumer satisfaction surveys in developing countries (for a recent review see [8]) there has been increasing concern in industrialized countries about the use of such instruments [9–12]. A number of the problems commonly identified can be at least partially resolved by careful design and implementation of the survey. Although several consumer satisfaction surveys have been designed and validated in industrialized country contexts, there are significant difficulties in transferring them directly to other situations where perceptions of quality may be very different. The research team reviewed available consumer satisfaction instruments and adapted them on the basis of existing Thai studies of consumer perceptions of quality of care, and consultation with Thai public health experts. The structure of the two survey instruments (one for outpatients and one for inpatients) is shown in Appendix 1. Doctors in Thai culture are traditionally held in very high regard, and thus people may be particularly unwilling to criticise them [13]. This factor is likely to be compounded by the Thai culture of Krengjai (consideration or respect) which inhibits open criticism of others. Combined, these factors suggested that Thai respondents to consumer satisfaction surveys may be particularly reluctant to express negative opinions about health care providers. Hence the researchers chose a combination of more and less direct questions. Direct questions on satisfaction used a Likert-type scale to assess aspects of physician care, nursing care, general cleanliness and amenities of the facility, and reception by other staff (such as registration clerks). Indirect questions did not ask about satisfaction per se but rather asked respondents to rate the clarity of explanations received from carers, and waiting time (for outpatients). The questionnaire concluded with open-ended questions about means through which to improve quality of care at the hospital. These open-ended questions provided an opportunity for respondents to voice concerns about dimensions of quality other than those already reviewed in the survey. In addition, on the inpatient questionnaire two ‘make or break’ questions were included asking whether (i) the patient him or herself would use the facility again and (ii) whether the patient would recommend friends or family to use the hospital. Both questionnaires were drafted in Thai, and pilot tested at two public hospitals (different from those where the survey was finally implemented) to ensure that they were easily comprehensible to patients.
Implementation method Nine hospitals in the Bangkok area including three public hospitals, three private for-profit hospitals and three private
User satisfaction in Bangkok
Table 1 Basic characteristics of sample hospitals Average length Hospital No. Ownership Bed No of stay (days) Occupancy rate ............................................................................................................................................................................................................................. 1 Public (BMA) 438 5.01 54.5 2 Public (MOPH) 428 5.39 80.6 3 Public (university) 832 8.74 72.7 4 Private for-profit (SET) 315 3.1 57.0 5 Private for-profit (SET) 184 3.2 69.4 6 Private for-profit 146 5.2 59.7 7 Private non-profit 171 4.3 59.5 8 Private non-profit 402 4.7 66.5 9 Private non-profit 312 4.3 70.6 BMA, Bangkok Metropolitan Authority; MOPH, Ministry of Public Health; SET, Stock Exchange of Thailand; LOS, Length of stay.
non-profit hospitals were purposively selected to make the nine hospitals broadly comparable in terms of bed numbers and scope of services offered (Table 1). A further factor in selecting hospitals was the hospital management’s willingness to participate in the research. The hospitals are referred to here by number in order to preserve anonymity. Within the public sector the three hospitals had different ownership: the Ministry of Public Health, the Bangkok Metropolitan Authority and the Ministry of Education (hospital 3 was a university teaching hospital and was considerably larger and had longer lengths of stay than any of the other facilities). For the inpatient satisfaction survey a four page selfadministered questionnaire with a pre-stamped envelope was given to 600 consecutive discharges (both adult and paediatric patients) at each hospital. Respondents were instructed by ward nurses to complete the form at home and return by mail within 2 weeks. This approach was used so that forms were completed in a neutral setting, where patients might be willing to be more critical. The questionnaires were distributed concurrently in the nine hospitals starting in October 1995. Parents were to act as proxy respondents for children aged under 14 years and assistance from relatives was required for patients who could not read or write. The questionnaire was anonymous; neither hospital identification number nor patient name was labelled on the form. For the outpatient satisfaction survey, 600, two-page questionnaires were randomly distributed to every third outpatient exiting obstetric/gynaecology, internal medicine, surgery and general medicine clinics of each hospital. Paediatric patients were excluded from the sample as it was a self-administered questionnaire. Patients were provided with a pencil and asked to fill the form before leaving the hospital, so for this survey the questionnaire was not completed in a neutral setting. The questionnaire was usually completed while waiting for drugs. Completed questionnaires were put into a closed box by respondents. The survey took 2 days to 2 weeks to administer depending on the caseload of a particular hospital. It was undertaken consecutively in each hospital during the period August–December 1995.
Profile of respondents The inpatient survey had a very low response rate: of the 5400 forms distributed just 1870 were returned, providing a response rate of 35%. There were significant differences in response rates between hospitals of different ownership with the highest response rate at public hospitals (45% response rate) and the lowest at private for-profit hospitals (26% response rate). The reasons for these differences are unclear. The response rate for the outpatient survey was significantly better as it was completed at the hospital: 3953 valid responses were received giving a response rate of 73.2%. Moreover, responses for this survey were relatively evenly distributed between hospitals, with the exception of one private forprofit hospital where some patients were mistakenly requested to mail back survey forms. Table 2 shows the profile of respondents in terms of their sex, age, education and social security cover. These data are compared with the same variables found during a 1 day, hospital bed census in each of the facilities. Thus for the inpatient survey it was possible to identify clearly the key characteristics of respondents. Unsurprisingly respondents were particularly those of working age (15–59 years), and they also tended to be better educated, having more postprimary school education. This finding was consistent across hospitals. Table 2 also suggests that women were considerably more likely than men to respond to the questionnaire. The data presented in Table 2 also suggest that SSS workers were over-represented in the sample, when compared to the bed census figures. There are at least two reasons why this may be the case: (i) SSS workers are aged 15–59 years and as the very young and the elderly were least likely to respond, age partially accounts for the over-representation of SSS workers; (ii) SSS workers may have shorter lengths of stay than other inpatients as the hospital is under pressure to save costs on their treatment. The bed census method is less likely to pick up patients with short lengths
311
V. Tangcharoensathien et al.
Table 2 Profile of respondents according to hospital type Private Private All hospitals Public for-profit non-profit ............................................................................................................................................................................................................................. Outpatient survey Female (%) 68 68 64 72 Aged < 15 years (%) 1 1 1 1 Aged 15–59 years (%) 93 89 95 94 Aged > 59 years (%) 6 9 4 5 With primary education only (%) 23 37 13 21 Covered by Social Security Scheme (%) 23 13 39 16 Inpatient survey Female (%) 67 75 50 71 Aged < 15 years (%) 4 4 4 5 Aged 15–59 years (%) 87 88 89 82 Aged > 59 years (%) 9 8 7 13 With primary education only (%) 39 58 28 25 Covered by Social Security Scheme (%) 18 15 34 10 Inpatient census Female (%) 56 55 50 62 Aged < 15 years (%) 16 18 9 17 Aged 15–59 years (%) 60 63 70 49 Aged > 59 years (%) 24 20 21 34 With primary education only (%) 46 53 44 30 Covered by Social Security Scheme (%) 7 5 22 3
of stay, whereas a survey administered to sequential discharges is more likely to pick up those with short lengths of stay. It is not uncommon for patients to have dual insurance cover and thus it is probable that a number of respondents mistakenly identified themselves as having SSS coverage whereas interviewer probing during the bed census found that care for this particular episode was actually being funded from an alternative source, such as an employers’ medical benefit scheme. Differences between the three groups of hospitals also emerge: the clientele at public hospitals tended to be less educated than that at private hospitals. This is probably explained by the strong correlation between income and education and the fact that lower income groups tend to prefer the cheaper services at government hospitals. The private for-profit hospitals have a larger number of male clients: this is probably partly due to the significant number of SSS workers at two of the private for-profit hospitals in the sample. Data presented in Table 2 are pooled by hospital ownership for ease of interpretation. During the analysis results for each hospital individually were examined as well as data pooled by hospital ownership. While v2 tests showed that differences between groups were often significant, there were some notable differences within hospital ownership groups. In particular hospital 3, the University teaching hospital, tended to be viewed quite differently from the other two public hospitals. In addition, hospital 4 was found to have quite a
312
different client profile from the other two private for-profit hospitals: it had no SSS workers registered with it and tended to draw its clients from more educated groups than hospitals 5 and 6.
Results Differences in satisfaction according to hospital ownership Table 3 shows the percentage of respondents stating that they received a ‘clear explanation’ from the doctor or relevant hospital staff member with respect to a number of different issues. For outpatient care a clear and highly significant pattern emerged: doctors at private for-profit hospitals were consistently thought to provide more clear explanations than doctors at public hospitals, and doctors at private non-profit hospitals in turn provided more clear explanations than doctors at for-profit ones. The only exception to this is explanation of drug use, where private for-profit pharmacies performed better than the private non-profit ones. For inpatient care a similar pattern emerged, although it was significant only for explanations about surgery and its consequences, and the nature of illness. In general, patients seeking care in public hospitals rated explanations received lower than patients seeking care in private for-profit hospitals, who in turn rated explanations lower than patients seeking care in private non-profit hospitals.
User satisfaction in Bangkok
Table 3 Percentage of respondents receiving a ‘clear explanation’ of findings and treatment Private Private All Public for-profit non-profit Explanation of hospitals hospitals hospitals hospitals P n ............................................................................................................................................................................................................................. Inpatient survey Lab and radiology findings 69.9 67.8 69.4 73.2 0.11 1634 Surgery and consequences 79.9 72.7 82.8 87.5 0.00 846 Treatment modalities 74.7 75.9 72.8 74.4 0.48 1616 Nature of illness 72.8 67.9 75.6 77.1 0.00 1700 Outpatient Survey Diagnosis 77.1 66.3 80.6 84.1 0.00 3701 Treatment 79.8 69.2 82.6 87.0 0.00 3626 Drug use 69.0 56.3 83.6 69.6 0.00 3546 Table 4 Percentage of inpatient respondents rating selected hospital characteristics as good or very good Private Private Public for-profit non-profit All Characteristic hospitals hospitals hospitals P hospitals n ............................................................................................................................................................................................................................. Cleanliness Comfort Availability of chairs Care taken by nurses Nurses’ manner Physical exam by doctor Clinical competency Regularity of doctor’s visit Doctor’s manner Overall quality
72.8 70.0 39.0 75.9 80.9 73.2 89.9 88.4 84.6 88.4
69.7 71.4 47.6 70.9 74.0 67.7 83.2 78.9 65.4 78.9
For most aspects of inpatient care, private for-profit hospitals received lower ratings than either of the other hospital groups (Table 4). The sole exception to this was on certain dimensions of amenities (such as comfort and availability of chairs). Private non-profit hospitals were rated most highly. On one dimension (the manner of physicians) public hospitals received a higher rating than private non-profit hospitals. This pattern was confirmed by the responses to the two ‘make or break’ questions (Table 5). Inpatient respondents Table 5 Percentage of inpatients stating they will use facility again or recommend to others, by hospital ownership Will Will use recommend facility to others again (%) n (%) n .......................................................................................................... Public hospitals 62 771 76 774 Private for-profit 55 448 59 449 Private non-profit 68 579 78 578 P 0.00 0.00 All 62 1798 73 1801
76.9 75.4 39.0 83.0 83.7 79.1 95.5 90.0 82.0 90.0
0.03 0.08 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
71.5 72.0 41.1 77.0 80.1 73.7 90.0 86.5 78.9 86.5
1824 1812 1796 1788 1780 1678 1778 1718 1702 1718
were asked (i) whether they would return to the same hospital if they needed inpatient care in the future and if they had alternative choices and (ii) whether they would recommend friends and relatives to use the hospital. Responses are consistent with the ratings respondents gave to aspects of quality in the three hospitals. Patients at non-profit private hospitals were most likely to return or recommend the hospital to a friend, and patients at private for-profit hospitals were least likely to do so. Responses differed considerably between inpatient and outpatient surveys. For the outpatient survey public hospitals received lower ratings than both groups of private hospitals for every aspect of quality of care (Table 6). On the whole, rankings of individual hospitals were consistent with the overall rankings of hospital by ownership; however there were a few exceptions to this. Most notably, the public teaching hospital was in all respects less well rated than the other two public hospitals for outpatient care, but respondents expressed a very high level of satisfaction with inpatient care received at the hospital. A further question included in the outpatient survey related to the length of waiting time for different types of services (to see a doctor, for registration, for drugs and for payment) and whether it was short, moderate or too long. Private
313
V. Tangcharoensathien et al.
Table 6 Percentage of outpatient respondents rating selected hospital characteristics as good or very good Private Private Public for-profit non-profit All Characteristic hospitals hospitals hospitals P hospitals n ............................................................................................................................................................................................................................. Comfort 48.3 70.0 77.5 0.00 65.6 3698 Cleanliness 57.2 77.4 85.4 0.00 74.0 3736 Queuing 45.0 57.3 61.7 0.00 55.4 3309 Doctor’s history taking 61.9 70.5 75.1 0.00 69.5 3614 Physical examination 65.3 70.7 80.7 0.00 73.0 3597 Time with doctor 53.9 58.6 67.2 0.00 60.6 3590 Doctor’s manner 77.2 83.2 88.8 0.00 83.4 3664 Care taken by nurse 55.5 73.4 78.8 0.00 69.7 3631 Nurses’ manner 65.7 76.3 81.4 0.00 74.9 3683 Overall 53.0 67.6 75.7 0.00 66.1 3490
Table 7 Differences in perceptions of quality between SSS and non-SSS groups in the outpatient survey (excluding hospital 6) SSS Non-SSS n P ............................................................................................................................................................................................................................. Percentage rating explanation as clear Explanation of diagnosis 64.3 72.3 1810 0.008 Explanation of treatment 65.9 75.2 1768 0.002 Explanation of drug use 66.0 66.4 1716 0.895 Percentage appraising aspects of doctor consultation as good or very good Attention to illness 59.8 66.7 1752 0.030 History taking 54.1 64.8 1753 0.001 Physical examination 59.3 67.0 1740 0.016 Time with doctor 44.0 56.8 1728 0.000 Doctor’s manner 77.8 79.0 1779 0.659
non-profit hospitals were most frequently thought to have acceptable waiting times, whereas there were most complaints about the waiting times at public hospitals. For example more than one-quarter of respondents thought that waiting time for registration at public hospitals was too long whereas the corresponding percentage for private non-profit hospitals was less than 10%.
Differences in level of satisfaction between SSS patients and other patients Interpretation of the evidence on payment method and perception of quality of care was complicated by inconsistent responses on insurance status. In particular a small percentage of respondents (approximately 10%) claimed that they were covered by the SSS for this visit, but were also being paid for by some other insurance scheme (e.g. CSMBS or private insurance). Some of these respondents may have been members of the SSS but had used a hospital with which they were not registered and therefore had to use alternative means to pay.
314
In the analysis presented here we simply compare those saying that they will use the SSS to pay for this visit and those who say they will not. As the two groups may be rather contaminated it is likely that our analysis underestimates the true difference in perceptions. In the outpatient survey, at hospitals 1, 2, 3 (the public hospitals) and 5 (a private for-profit hospital) 10–20% of respondents had SSS coverage. At hospital 6 nearly 70% of respondents had SSS coverage. The other hospitals either accepted no or very few SSS patients. Initially the ratings relating to the physician consultation given by SSS patients in hospitals 1, 2, 3, 5 and 6, where there were a substantial number of SSS patients, were compared with those given by non-SSS patients. In each hospital, with the exception of hospital 6, higher ratings had been awarded by non-SSS patients than by SSS patients. It is possible that physicians at hospital 6 dealt with so many SSS patients that it was difficult for them to discriminate against them. Table 7 summarizes the results for hospitals 1, 2, 3 and 5: large and significant differences emerge between SSS and non-SSS patients on virtually all of the various characteristics examined. In particular, SSS patients were
User satisfaction in Bangkok
considerably more likely than non-SSS patients to think that the time which the doctor spent with them was insufficient or that the doctor did not take a full history of the condition. The small percentage of patients at government hospitals who were exempted through the low income scheme or carried a voluntary government health insurance card (approximately 3% in total), would not have paid for care on a fee-for-service basis. Furthermore, it is possible that patients paying out-of-pocket (and hence constrained by their own ability to pay) may not have been viewed by physicians as being as remunerative as insured patients paying on a feefor-service basis. To ensure that the comparison captured only and strictly difference in payment mechanism, a further analysis on a hospital-by-hospital basis compared (i) ratings awarded by SSS patients versus CSMBS patients in public hospitals and (ii) ratings awarded by SSS patients versus those with private insurance or employer insurance schemes in private hospital 5. The results showed differences in the ratings awarded by SSS and insured fee-for-service patients of a similar or greater magnitude than those reported in Table 7, however fewer results were significant, perhaps due to smaller sample size. A similar analysis was carried out for the inpatient satisfaction survey, however no consistent pattern between ratings of SSS and non-SSS groups emerged and none of the differences observed were found to be significant. This was the case for each individual hospital and for the whole sample.
Discussion and conclusions The survey illustrated some interesting differences in consumer perceptions of quality of care and how they varied between hospitals of different ownership and by patient’s payment status. For both inpatient and outpatient care the private non-profit hospitals were highly appraised, but whereas public hospitals were generally better thought of than private for-profit hospitals for inpatient care, the reverse was true for outpatient care. The only dimensions in which private for-profit hospitals out-performed public hospitals for inpatient care was with respect to the amenities available, such as comfort of surroundings, availability of chairs etc. These findings tend to confirm the popular stereotype suggested in the introduction: private for-profit hospitals perform better with respect to hotel aspects of care and convenience (as reflected in waiting time data), but for more serious conditions, which require admission, patients prefer the services provided at public hospitals. Because of the problems encountered in distinguishing between those who were and were not covered by the SSS for this particular episode of care, our conclusions about payment status and consumer satisfaction with quality remain tentative. The results found here suggest that further investigation of this issue might be valuable. At the time of the study, in all study hospitals SSS patients were seen in the same clinics as regular patients and thus it is unlikely that there were substantial differences in certain aspects of care such as the comfort and cleanliness of
surroundings. Hence our analysis focused on the physician consultation where it is possible that SSS patients received less attention than fee-paying patients. Interviews with hospital staff suggested that over time hospitals were developing ways to minimize the costs of treating SSS patients. These mechanisms may have negative effects upon quality of care. At the time of the survey only outpatient care appears to have been negatively affected, but stronger hospital responses to the incentives inherent in capitation payment are rapidly evolving. Satisfaction with health care is closely linked to expectations and the degree to which expectations are met. A clearer understanding of people’s expectations may help with interpretation of the results. For example, there has been substantial press coverage of cases where SSS workers have been denied access to emergency care. Discussion in the media may have given SSS patients the impression that they were receiving ‘second class’ care and this could in turn have influenced their responses to the questions. Hence it is possible that the observed differences in ratings between SSS and non-SSS groups are due to SSS workers taking a more critical attitude towards services received. In order to investigate this possibility, ratings between SSS and non-SSS groups were compared for dimensions such as comfort and cleanliness which would not differ across the two groups. For these variables there was no clear pattern in ratings and none of the observed differences were significant. Another example of the importance of expectations relates to the university teaching hospital which was a clear outlier in terms of consumer satisfaction with outpatient care provided; it performed significantly less well than either of the other government hospitals. Although there are problems at the outpatient departments of the hospital, notably a very large patient caseload in the region of 4000 outpatient visits per day, the poor evaluations of care by users may also be partially due to high expectations. The university teaching hospital in question is well known within Thailand and thus people may have been particularly shocked by the impression they took away of the outpatient departments. Although the consumer survey provided interesting insights it is clearly an imperfect instrument and problems were encountered by the researchers in implementing the survey. Key shortcomings of the study were: (i) as commonly found elsewhere a relatively high proportion of respondents (approximately 75%) appeared satisfied with the care received (rating various aspects of quality as good or very good); (ii) very low response rates amongst certain identifiable groups, particularly the elderly, the less educated, and children, were also problematic. The response rate was particularly low for the inpatient survey where respondents were asked to mail back questionnaires. No attempt was made by the researchers to correct for non-respondent bias as the only data available on the typical profile of hospital users came from a 1 day bed census which might in itself not be very reliable. It is probable that the groups which did not respond
315
V. Tangcharoensathien et al.
are those who would be least likely to articulate dissatisfaction with care received. (iii) the survey was implemented in only nine hospitals; Bangkok has a total of over 150 hospitals. The small sample size means that the data on differences between groups of hospitals should be interpreted with caution. The fact that bed occupancy rates also varied within groups, and that this may have an independent effect on patient satisfaction (via availability of staff ), should also be taken into account.
TS3-CT94–0325 (DG12 HSMU). The researchers based in the Health Economics and Financing Programme, London School of Hygiene and Tropical Medicine are further supported by the Department for International Development, Great Britain. We thank the staff and management of the nine hospitals in Bangkok who permitted us to carry out this survey and assisted with its implementation.
Qualitative approaches are needed to supplement the survey results and provide more in-depth analysis of expectations, true levels of satisfaction and the perspectives of more vulnerable groups. The current set of health policies in Thailand point to a future where an increasing proportion of the population have their health care funded by social health insurance schemes, and are able to seek care from public or private providers. The study reported here showed that among public sector providers there were problems related to lack of responsiveness to consumers, whereas in the private sector concerns focused more upon clinical standards of care. Simultaneous to developing the necessary structures to expand insurance coverage in Thailand, it is important to put in place adequate systems for monitoring quality of care. These systems will include both top down monitoring plus incentives and tools for hospitals to evaluate their own quality. Hospital managers, as well as policy makers, found the results from the study useful. Although some of the hospitals (mainly the non-profit facilities) had implemented some type of consumer survey previously, this was often in an ad hoc fashion. This was thus the first time that any of the hospital managers had had the opportunity to compare users’ perceptions of their hospital with those of other hospitals. Hospital managers were understandably most interested in those areas where management interventions could relatively easily improve consumers’ perceptions of quality of care such as when patients perceived bottle-necks in flow of services (leading to long waiting times for a particular service) or gave poor ratings for specific aspects of care (such as the availability of chairs for visitors). Researchers at the Health Services Research Institute have revised the survey form to focus on questions which were found to be particularly useful. This revised form was then distributed to all public and private hospitals in Thailand, together with instructions about how to implement the survey, desired sample size, dummy tables to use for analysis and a summary presentation of the results from the nine hospitals already studied. These are small measures, but they form important first steps in constructing the necessary quality monitoring mechanisms in Thailand.
1. The Economist Intelligence Unit. Profile – hospital care in Thailand. Healthcare industry: Healthcare Asia 1998; 2nd Quarter: 17–34.
References
2. Nittayaramphong S, Tangcharoensathien V. Thailand: Private health care out of control? Health Policy Planning 1994; 9: 31–40. 3. Brugha R, Zwi A. Improving the quality of private sector delivery of public health services: challenges and strategies. Health Policy Planning 1998; 13: 107–120. 4. Bennett S. Private health care and public policy objectives. In Colclough C, ed., Education and Health in Developing Countries: Miracle or Mirage? Oxford: Clarendon Press, 1997: pp. 93–123. 5. Yu S, Anderson GF. Achieving universal health insurance in Korea: A model for other developing countries? Health Policy 1992; 20: 289–299. 6. Peabody JW, Lee S, Bickel SR. Health for all in the Republic of Korea: one country’s experience with implementing universal health care. Health Policy 1995; 31: 29–42. 7. Peabody JW, Yu JC, Wang Y, Bickel SR. Health system reform in the Republic of China: Formulating policy in a market-based health system. J Am Med Assoc 1995; 273: 777–781. 8. Haddad S, Fournier P, Potvin L. Measuring lay people’s perceptions of the quality of primary health care services in developing countries: Validation of a 20-item scale. Int J Qual Health Care 1998; 10: 93–104. 9. Avis M, Bond M, Arthur A. Questioning patient satisfaction: an empirical investigation in two outpatient clinics. Soc Sci Med 1997; 44: 85–92. 10. Owens DJ, Batchelor C. Patient satisfaction and the elderly. Soc Sci Med 1996; 42: 1483–1491. 11. Calnan M. Lay evaluation of medicine and medical practice: report of a pilot study. Int J Health Serv 1988; 18: 311–322. 12. Carr-Hill R. The measurement of patient satisfaction. J Public Health Med 1992; 14: 236–249. 13. Smith HE. Doctors and society: A Northern Thai Study. Soc Sci Med 1982; 16: 515–526.
Acknowledgements The research study was supported financially by the European Commission, DGXII under fixed contribution contract No
316
Accepted for publication 3 March 1999
User satisfaction in Bangkok
Appendix 1 Fields included in the questionnaires Outpatient questionnaire
Inpatient questionnaire 1. Patient characteristics – as for outpatient survey. 2. Reasons for choice of hospital, diagnosis/condition.
1. Patient characteristics: area of residence, sex, education, insurance coverage for this visit, previous use of hospital.
3. Quality of explanations by physician (on diagnosis, nature of illness, treatment modalities, surgery and consequences, lab and radiology findings, extent to which decision on treatments was made jointly by doctor and patient).
2. Reasons for choice of hospital, diagnosis/condition for which care was sought.
4. Knowledge of name of attending physician.
3. Quality of explanations by physician and pharmacist.
5. Price charged and perception of price.
4. Perception of waiting time at various service points (registration, consultation with doctor, cashier and dispensary).
6. Length of stay and perception as to how appropriate.
5. Price charged and perception of price.
7. Five-point Likert scale rating of clinical and non-clinical aspects of quality of care.
6. Five-point Likert scale rating of quality of care for clinical (doctor and nursing services) and non-clinical (cleanliness, amenities) aspects of care.
8. ‘Make or break’ questions on whether the patient would return to the same hospital if requiring further hospitalization or would recommend friends or relatives to use this hospital.
7. Open-ended question: three suggestions to improve quality of care.
9. Open-ended question: three suggestions to improve quality of care.
317
