Background
Apart from the inherent goal of health promotion for the population, a responsive health system should commit itself to meet the psychological need of the people it serves and to undertake an examination of what people care about when they interact with the health system [
1]. Improving these non-health functions of a health system is important because it is an inalienable component to increasing people's well-being, that being a universal and ultimate mission of a health system. Appropriately, the concept of responsiveness was introduced and a related questionnaire was also designed by the World Health Organization (WHO) to measure how well a health system meets the legitimate expectations of the population for the non-health enhancing aspects [
2,
3]. To emphasize its significance, the World Health Report 2000 labeled assurance of responsiveness as one of three intrinsic standards (in addition to health improvement and financial fairness) to evaluate health system performances [
4].
Through the extensive processes of literature review and panel discussions, two elements were defined by WHO to measure the concept of responsiveness: respect for persons and client orientation [
5]. Respect for persons, intending to capture ethical aspects of the interaction of individuals with the health system, includes three sub-elements: dignity, autonomy and confidentiality [
6‐
8]. On the other hand, client orientation (which mainly gauges the components of consumer satisfaction) has four sub-elements: prompt attention, quality of basic amenities, access to social supports for hospitalized individuals, and choice of health providers [
9,
10]. Using these seven sub-elements as the framework, WHO developed a structured questionnaire to measure responsiveness [
11]. In 1999, WHO conducted key-informant surveys in 35 countries to collect data on responsiveness and then used regression models to assess responsive level for the other nations that did not implement such a survey [
11,
12]. Finally, WHO published a global ranking of responsiveness in the World Health Report 2000.
For the intervening years since its publication, the report has generated much discussion and criticism. Some critics questioned the validity in using a single composite score to compare global health systems [
13,
14]. Others argued that using the prevailing "Western wisdom" of key informants (experts) to assess responsiveness might be less accurate than the judgments by patients themselves [
15,
16]. Blendon analyzed data from 17 European countries to demonstrate significant differences between the national ranking of WHO defined responsiveness and the satisfaction of the general public who actually experienced these health systems [
17]. The report's publication also spawned contentious debates on the appropriateness of undertaking international comparisons of health system performance, given that the platforms of health systems may be quite diverse in societies with different cultural, economical or political backgrounds [
18,
19].
Despite these issues and debates, there has not been, to our knowledge, any empirical research undertaken studying the suitability of the seven sub-elements used by WHO as the national standard to define responsiveness for a health system. The main objective of this study, then, was to assess if the proposed seven sub-elements are applicable in Asian countries such as Taiwan. If not, we would seek to determine what components should be reorganized or what other constituents should be added in order to formulate a sounder indicator for the evaluation of responsiveness in Taiwan.
Results
Table
1 summarizes the basic demographic characteristics of the survey respondents. About half (48.5%) of them were scholars, one-fourth were clinicians, and another one-fourth (26.6%) were from governmental agencies. Of respondents, males dominated (71.2%), and about half were aged below 50 years (45.5%), while those above 60 years accounted for 18.9%.
Table 1
The demographic characteristics of surveyed key informants.
Total | 132 | (100.0) |
Employment status | | |
Health Related Governmental Agencies* | 20 | (15.2) |
Other Governmental Agencies†
| 15 | (11.4) |
Universities and Research Institutes | 64 | (48.5) |
Hospitals and Medical Organizations | 33 | (25.0) |
Gender | | |
Male | 94 | (71.2) |
Female | 38 | (28.8) |
Age | | |
20–49 | 60 | (45.5) |
50–59 | 47 | (35.6) |
60+ | 25 | (18.9) |
Table
2 displays the differences of factor structures constructed by the WHO study group (WHO model) and by this study (Taiwan model). In the Taiwan model, each of the retained five factors displayed eigenvalues greater than 1. Combined, these five factors accounted for 63.5% of the total variance. Using factor loading 0.4 as the cutoff criterion, 8, 9, 8, 8, and 4 items were found to load on the first (Respect), the second (Access), the third (Basic Amenities), the fourth (Confidentiality), and the fifth (Social Support) factors, respectively. The Cronbach alphas of five identified factors were between 0.79 and 0.89, indicating that their internal consistency was highly reliable.
Table 2
Principal component analysis to identify components of responsiveness in Taiwan.
| Eigenvalue | 10.675 | 4.518 | 3.416 | 1.949 | 1.655 |
| Cronbach's alpha* | 0.891 | 0.811 | 0.856 | 0.876 | 0.788 |
Dignity
| | | | | | |
1 | Treated with respect |
0.530
| 0.149 | 0.030 | 0.251 | 0.032 |
2 | Common human rights | 0.351 | 0.033 | 0.010 |
0.483
| -0.045 |
3 | Free to discuss concerns |
0.816
| -0.027 | 0.063 | -0.041 | 0.081 |
4 | Encouraged to ask questions |
0.831
| 0.012 | 0.115 | -0.046 | 0.013 |
5 | Respect privacy in physical examination |
0.661
| -0.099 | -0.059 | 0.325 | 0.033 |
5A | Overall score |
0.465
| 0.132 | 0.122 |
0.445
| -0.128 |
Autonomy
| | | | | | |
6 | Provided with information |
0.830
| -0.016 | -0.094 | 0.046 | 0.021 |
7 | Consult on preferences |
0.835
| -0.083 | 0.033 | 0.002 | 0.059 |
8 | Patient consent | 0.257 | 0.096 | -0.162 |
0.644
| -0.131 |
8A | Overall score |
0.529
| 0.042 | 0.037 |
0.437
| -0.095 |
Confidentiality
| | | | | | |
9 | Protect confidentiality | 0.369 | -0.011 | -0.106 |
0.624
| -0.050 |
10 | Confidentiality of patient information | 0.090 | -0.131 | 0.004 |
0.881
| 0.087 |
11 | Confidentiality of medical records | -0.077 | -0.093 | 0.059 |
0.816
| 0.158 |
11A | Overall score | 0.077 | 0.042 | 0.063 |
0.831
| 0.049 |
Prompt attention
| | | | | | |
12 | Geographic access | -0.065 |
0.739
| -0.106 | -0.093 | -0.037 |
13 | Fast access to emergency care | -0.035 |
0.711
| -0.130 | 0.059 | -0.004 |
14 | Reasonable waiting time | -0.042 |
0.603
| 0.352 | -0.034 | -0.166 |
15 | Reasonable waiting time for non-emergency surgery | -0.091 |
0.756
| -0.016 | 0.032 | 0.087 |
15A | Overall score | -0.188 |
0.836
| 0.029 | 0.253 | -0.005 |
Social support
| | | | | | |
16 | Allow visitors | 0.088 | 0.023 | 0.010 | -0.032 |
0.829
|
17 | Take care of personal needs | 0.146 | 0.107 | 0.001 | -0.087 |
0.814
|
18 | Practice religious acts | -0.095 | -0.062 | 0.036 | 0.151 |
0.735
|
18A | Overall score | -0.199 | 0.355 | 0.045 | 0.289 |
0.541
|
Basic amenities
| | | | | | |
19 | Cleanliness of unit | 0.105 | -0.160 |
0.752
| -0.060 | 0.139 |
20 | Maintenance of building | -0.115 | -0.167 |
0.801
| 0.132 | 0.081 |
21 | Adequacy of furniture | 0.030 | -0.164 |
0.720
| 0.043 | -0.072 |
22 | Nutrition and edibility | 0.029 | 0.389 |
0.466
| -0.016 | -0.231 |
23 | Clean water | -0.002 | 0.207 |
0.666
| -0.111 | 0.040 |
24 | Cleanliness of toilets | 0.069 | -0.093 |
0.794
| -0.099 | 0.095 |
25 | Cleanliness of linen | 0.125 | 0.171 |
0.656
| -0.094 | -0.064 |
25A | Overall score | -0.114 | 0.079 |
0.811
| 0.219 | -0.018 |
Choice of providers
| | | | | | |
26 | Choice of provider at health care unit | 0.231 |
0.630
| 0.074 | -0.134 | -0.100 |
27 | Choice between units | 0.147 |
0.552
| 0.009 | -0.093 | 0.391 |
28 | Choose to see specialist | 0.122 |
0.666
| -0.116 | -0.185 | 0.222 |
28A | Overall score | 0.072 |
0.783
| 0.015 | 0.080 | 0.122 |
The item contents of two factors (Basic Amenities and Social Support) were unchanged between the WHO and Taiwan models. The factor Autonomy defined in the WHO model no longer existed in the Taiwan model because it contributed two items to the Respect factor and also contributed the last remaining item to the Confidentiality factor of the Taiwan model. The factor Access in the Taiwan model essentially comprised the combination of the Prompt Attention and Choice of Provider factors of the WHO model. Most items of the Dignity factor of the WHO model were retained in the Respect factor of the Taiwan model, with the exception of one item related to the human rights issue, which was reallocated to the Confidentiality factor of the Taiwan model.
In Table
3, we used data collected in this study to delineate the relationship between the overall responsiveness score and five constructed factors in the Taiwan model. After adjusting for demographic variations, regression statistics of this model fit very well (R
2 = 0.70). Except for Social Support, the other four factors (Respect, Access, Confidentiality, and Basic Amenities) identified in the Taiwan model could significantly explain the variance of the dependent variable (overall responsiveness score).
Table 3
The relationship between total responsiveness score and the five extracted factors†
Intercept | 0.574 (0.304) |
Respect | 0.203 (0.079)* |
Confidentiality | 0.229 (0.082)** |
Access | 0.351 (0.066)*** |
Social Support | 0.075 (0.060) |
Basic Amenities | 0.313 (0.064)*** |
R2
| 0.700 |
F-value | 27.37 |
P-value | <0.0001 |
After thoroughly interactive discussions, the focus-group participants regarded dignity, prompt attention and confidentiality as three most important components a responsive health system should embody. The results of the focus group interviews also indicated that medical ethics of health providers, and communication between patients and physicians were two additional areas that are priorities for Taiwanese citizens in judging the responsiveness of a health system.
Medical ethics (e.g., providers' honest behaviors and treating patients without discrimination) were frequently brought up during the group interviews. The issues that were of the most concern were discrimination against the poor and the alleged access to better health care via "red envelope" bribery. Concerning communication, most participants mentioned that in a responsive health system health providers should listen to patients carefully, explain medical terms in language that is understandable, and spend more time with patients in order to answer their questions.
Discussion
This study has several inherent limitations. First, to our knowledge, a globally accepted assessment framework for responsiveness has not been theoretically formulated; therefore, this study just represents empirical research designed to test applicability of the WHO proposed concept in an Asian country and also to explore prospective dimensions in delineating responsiveness of the health system in Taiwan. Second, although the nominated key-informants may better comprehend achievements of a health system on the elements of responsiveness, participants of the key informant survey, as frequently criticized [
17], may be biased by theoretical prejudgment. The consensus of key informants may not be able to reflect opinions of the general public, the real users of a health system. Third, participants of focus groups were volunteers who may have a vested interest in health system issues. The major reason for us to conduct a key-informant survey, instead of doing a national representative survey, was to increase comparability with the WHO estimated factor structure [
11]. The seven sub-elements of the WHO proposed responsiveness measure were constructed by using key-informant surveys from 35 countries. We followed the standard protocol developed by WHO [
11] to conduct the key-informant survey in order to do an item-by-item comparison in principal component analysis between factors constructed in this study and in the WHO proposed model. While bearing in mind the possible limitation of a key-informant survey, we conducted this study to do a model-building exploration. To fulfill this purpose, we also held three focus-group interviews to better understand, from the user's perspectives, embedded meanings behind the factor structure extracted in this study and also probe for new dimensions for the Taiwanese framework of responsiveness.
This Taiwanese empirical study indicates seven WHO defined sub-elements may have to be reorganized to fit Taiwanese society. Using 0.4 as the factor loading cutoff point to determine whether or not the item should be considered a good factor indicator, two original sub-element scores (Dignity and Autonomy) loaded onto two factors (Respect and Confidentiality), rather than being retained in a unique factor. On the other hand, for the newly defined factors, three out of five contained more than one original sub-element scores (Table
2).
In the element "respect for persons", the sub-element Autonomy seems not conceptualized by Taiwanese respondents as a unique domain, of which the items related to alternative treatment options (Q6 and Q7 in Table
2) were categorized into the sub-element Respect, while the item related to patient consent (Q8) was categorized into the sub-element Confidentiality. Although patient rights have recently been advocated in Taiwan, it is a tradition that the opinions of a patient's key family members (e.g., father, husband, or the oldest brother) usually determine the patient's treatment. In focus-group discussions, the participants frequently expressed the same experience: when the diagnosis of a cancer is confirmed by the pathologist, it is usually a practice routine for a physician in Taiwan to obtain authorization for treatment from the patient's key family members before informing the patient of his/her true condition. Generally speaking, this study shows less evidence of a conceptual difference between dignity and autonomy among Taiwan respondents than was found in the WHO study. It may stem from Chinese paternalistic culture that the concept of patient autonomy is not currently deemed as the "legitimate expectation" to a health system for the majority of Taiwanese people.
For the element "client orientation", the items designated to measure "prompt attention" and "choice of health providers" perfectly merged (so we named it "Access" in Table
2), indicating that unrestricted accessibility of the health system in Taiwan may pave a common trait for these two sub-elements. Prompt attention (mainly implying minimization of waiting time and waiting list) and free choice of health providers are two major achievements of national health insurance (NHI) in Taiwan. This Taiwanese compulsory health insurance, initiated in 1995, covered 97.1% of total population in 2002 [
20]. Under this comprehensive coverage, there is no restriction for patients to choose physicians or medical institutes. Moreover, no referral is necessary from gate-keepers (e.g., family physicians) to visit specialists. Patients can also see any physician they prefer without incurring an extra charge. As a result, choice of health providers is not a barrier when accessing Taiwan's health services. On the other hand, the main payment mechanism in NHI is a "fee for service," which provides an incentive for physicians and medical institutes to enroll as many patients as their capacities enable; thus, in Taiwan the problems of long waiting time and waiting list are usually resolved by the system itself.
The results of the focus group discussions emphasized patient-physician communication and medical ethics to exert profound influence on the responsiveness of a health system. The communication component has already been adopted in WHO household survey for responsiveness [
21]. It indicates a global consensus in improving patient-physician communication to raise the responsiveness level of a health system. On the other hand, the Taiwanese focus group participants particularly underscored the oriental viewpoints of the fundamental value of responsiveness in relation to high-standard medical ethics to treat patients fairly and empathetically. In building a measurement index for responsiveness in Taiwan, it could be suggested that the factor Communication be added to the element "client orientation," and the factor Medical Ethics be appended to the element "respect for person."
Table
3 reveals that the re-organized five sub-elements (Taiwan model) could explain 70% variability of the overall responsiveness score, the respondents' general impression to the non-health enhancing performance of the health system in Taiwan. This satisfactory result means that an equally accurate representation of responsiveness in Taiwan can be made using a revised WHO composite index (five sub-element scores) identified in this empirical study. Accompanied with two aforementioned components suggested by focus-group participants, we may thus propose a modified structure to define dimensions of responsiveness from Taiwanese perspectives. The element "respect for persons" should consist of three important factors (Respect, Confidentiality and Medical Ethics). However, the element "client orientation" should include four essential components (Access, Communication, Social Support and Basic Amenities). Of course, this proposed model requires further confirmation to validate its appropriateness in evaluating responsiveness in Taiwan. A related questionnaire has been revised and developed accordingly, and a national representative survey is also planned to carry out to solicit user view on responsiveness.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CCH participated in concept development, statistical analysis, and paper writing. LC and YWH both participated in questionnaire development and focus group discussion. WY participated in concept development and study design. CCS participated in questionnaire development and statistical analysis.