Health System Responsiveness and Its Associated Factors Among Outpatients in Primary Health Care Facilities, Asagirt District, NorthShewa Zone, Ethiopia, 2021: Cross Sectional Study Design

18 Background: Health system responsiveness is defined as the outcome of designing health facility 19 relationships in such a way that they are familiar and respond appropriately to patients’ universally 20 legitimate expectations. Even though different strategies have been implemented to measure 21 responsiveness, only scanty evidence exists in Sub-Saharan Africa. In Ethiopia information about 22 the level of health system responsiveness among outpatients is scant. Assessing responsiveness 23 could help facilities in improving service delivery based on patient expectations. Objective: The study aimed to assess health system responsiveness and associated factors among 25 outpatients in primary health care facilities, Asagirt District, North Shewa Zone, Ethiopia, 2021. 26 Methods: Facility-based cross-sectional quantitative study was implemented between 30 th March 27 and April 30/2021. A systematic random sampling technique was employed to select 423 28 participants, and interviewer-administered data were collected using a structured and pretested 29 questionnaires. Both bivariable and multivariable logistic regressions were employed to identify 2 Results: The overall health system responsiveness was 66.2% (95% CI: 61.4% - 70.7%). Confidentiality and dignity domains were the highest responsiveness score. Health system 35 responsiveness was higher among satisfied outpatients (AOR: 9.9, 95% CI: 5.11-19.46), utilized 36 private clinics (AOR: 8.8, 95% CI: 4.32-18.25), and no transport cost (AOR: 1.7, 95% CI: 1.03- 37 2.92) in the study setting. 38 Conclusion: Overall health system responsiveness was higher as compared to other case-specific 39 study in Ethiopia. The domains of Autonomy, Waiting time, Basic amenities, and Choice were 40 identified as vital areas needing the effort to raise responsiveness of health care service in the 41 District. HSR was higher in private than public healthcare facilities, among satisfied clients and 42 those who didn’t pay for transport on their way to the health facility than their counterparts. Thus, 43 enhancing patient satisfaction, using input from service users, Collaboration, and experience 44 exchange between public and private facilities will be important interventions to improve HSR. 45

All health systems are expected to achieve the goals of good health, responsiveness to the 48 expectations of the population, and fairness of financial contribution [1][2][3][4]. From these goals; 49 health system responsiveness (HSR) is defined by the World Health Organization (WHO) as "how 50 well the health system meets the legitimate expectations of the population for the non-health 51 enhancing aspects of the health system" [5]. Health systems can be evaluated as a whole in any 52 type of interaction by summarizing into responsiveness [6,7]. The concept entails the experience 53 of people's fundamental interaction and different factors shaping their interaction with the health 54 system. This intern can help to anticipate and adapt to the existing and future health needs for a 55 better health outcome [1,5,8]. To provide appropriate and efficient care delivery, more responsive 56 and updated health systems with giving attention to intrinsic values and safeguarding of the rights 57 of patients are needed [2,4,[9][10][11][12]. However, the burden of diseases and conflicts in low and 58 middle-income countries threatening the capacity of health systems to respond to the population 59 they serve [13][14][15]. But the fulfillment of patient expectations is more important than other factors 60 for a better health outcome [16]. If health system responsiveness has improved other associated 61 health outcomes improved as well [7]. 62 Responsiveness has been operationalized into eight domains as respect for the dignity of persons; 63 autonomy to participate in health-related decisions; confidentiality; prompt attention; adequate 64 quality of care; communication; access to social support networks; and choice of health care 65 providers [1,2,8,17]. 66 Despite challenges for measuring responsiveness, additional refinement of strategy and consistent 67 monitoring are needed to achieve its goal [5,14,18]. Notably, those low and middle-income 68 countries are needed to give attention to equity health access at local and global aspects [19][20][21]. 69 Studying health system responsiveness is needed to improve patient experiences and their 70 satisfaction in the sphere of non-medical aspects [22][23][24][25]. A patient-centered and acceptable quality 71 across the continuum of care is essential through considering social norms, relationships, values, 72 and trust within societies [26]. The measurement of health system responsiveness also will help to 73 identify the level of performance health facilities [1]. 74 For a better and comprehensive understanding of non-health enhancing aspects of health systems, 75 measuring health care responsiveness is necessary [1,14]. This is because the fulfillment of 76 patients' expectation is more important than other factors for a better health outcome administrative unit). Asagirt District is located in North Shewa Zone, Amhara National Regional 100 State of the eastern edge of Ethiopia. The 2020 projected population of the District was 57,320. 101 Of whom 30,240 were males. The District has a total of 20 functional health facilities: 3 public 102 health centers, 2 primary private clinics, and 15 health posts (community-level health facilities 103 providing basic preventive and medical care). In 2021 a total of 52 health professionals and 23 104 health extension workers were served the District. According to the District health managers' 105 report, there was an average of one thousand seven hundred (1700) patients visiting health centers 106 and private clinics within a month. 107

108
A facility-based cross-sectional quantitative study design was conducted to assess health system 109 responsiveness among outpatients from 30 th March to April 30 /2021. 110

111
All outpatients who received health care services in primary health care facilities residing in 112 Asagirt District constitute the source population of this study. Patients who received health care 113 services as an outpatient in the selected primary health care facilities were included in the study. 114 Whereas patients whose age below 18 years, and all outpatients who were utilized health posts 115 were excluded from this study.  Then at every K th interval (K = N/n) where N= total clients who was received health care services 132 within the study period n= required sample size, thus K=1700/423=4. Then, the first patient was 133 randomly identified from 4 by lottery method, and then every 4 th patient was taken into the study 134 till the required number of study participants for each facility in the outpatient department was Autonomy (4). The eighth domain (access to social support network) was not assessed since it is 144 used for assessing inpatients (hospitalization) only [11,31]. All the 28 items were computed and 145 then it was dichotomized as "acceptable" and "unacceptable" by the demarcation threshold 146 formula as: [37-39]. Accordingly, 147 those who scored 73 and above HSR was considered as "Acceptable" and below considered as 148 "Unacceptable". 149 Likewise, all the seven domains were added separately and grouped as good and poor by the above 150 formula [37-39]. Above the cut-off point to determine "Good" performance, while including cutoff 151 point and below scores were considered as "Poor" for each domain independently. 152

Perceived satisfaction of clients: Patient satisfaction was measured by using 5 questions on a 153
five-point Likert scale with five response categories (1 'very dissatisfied' to 5 'very satisfied'), 154 and finally it was grouped by using the demarcation threshold formula [38,39]. And those who 155 scored 15 and above were considered as "Satisfied" whereas below 15 was considered as 156 "Dissatisfied". 157 Perceived quality of care score: assessed by 12 questions of the clients' perception about the 158 services they offered, professionalism of provider as well as, the patient values and interests in the 159 services. Then it was dichotomized into "high" for those who scored above 37 and "low" for those 160 who scored 37 and less [37]. 161

Out of pocket payment; was assessed by Yes/ No question [23]
. 162 PHQ-9 : was assessed by 9 depression questions to assess whether the patient has depression or 163 not ranging from 1 'always' to 4 'not at all' after which it was dichotomized as "poor" and "good" 164 with a cutoff point of 23 [40]. 165

Data collection tools and procedure
166 Closed-ended interview questionnaires adapted from WHO health system responsiveness and 167 questionnaires developed from reviewing different related literatures, were used for data 168 collection. The questionnaire was prepared in English first, then translated to Amharic (local 169 language), and then retranslated back to the English language to check its consistency. The 170 reliability of the tools was checked by Cronbach's alpha reliability test. Accordingly, values for 171 PHQ-9 (0.87), for satisfaction (0.89), for perceived quality of health care (0.96), and average 172 Cronbach's alpha for all domains was 0.92, all showed high reliability above the required cut-off 173 0.70. The questionnaire mainly includes socio-demographic assessment, health facility-related, 174 WHO responsiveness assessment questionnaires, perception on quality of care, and health 175 insurance membership. The data collectors went and collected the data from participants' after 176 they have received the services on their way to the home (exit interview). The data were collected 177 daily. A data collector has approached by introducing him/her self and interviewed the selected 178 respondent after informed consent was obtained. 179 Data quality assurance 180 Before the data collection, one-day training was given for all data collectors and supervisors by 181 the principal investigator about the mechanism of data collection to have a similar understanding. 182 Five B.Sc. Health Officers for data collectors and two supervisors of the same field who were 183 working out of study areas participated in the data collection process. The training process focused 184 mainly on the objective of the study, how to ask and fill the questionnaires, selection criteria of 185 patients and how to approach the respondents without introducing biases. Additionally, the facility 186 workers were not allowed either to see or hear the patients' response. During the data collection, 187 data collectors were assigned for the supervisor for better monitoring. Before starting the actual 188 data collection, the data collectors had practiced in the field and the questionnaires were pretested 189 on 21 (5%) patients in the nearby District (Angolela and Tera District). The data collectors and the 190 principal investigator had assessed the clarity and completeness of the questionnaires. Findings 191 and experiences from the pretest were utilized in modifying the data collection tool. When there 192 was any problem during the data collection process, the investigator had discussed it with the 193 supervisor and a solution was given on a daily bases. 194

195
Once the collected data were checked for completeness then the data were entered into the Epi-196 data version 4.6 Software Package. Then it was exported to Stata version 14 statistical software 197 packages for cleaning, coding, and analysis. A two-stage data analysis (descriptive and inferential) 198 was conducted. The descriptive statistics were described using frequency, percentage, mean and 199 standard deviation and presented by a figure, table, and text. All continuous independent variables 200 were categorized. Normality tests such as kurtosis and skewness were employed to identify which 201 summary measure is appropriate to use. Multicollinearity among independent variables were 202 checked using variance inflation factor (VIF) and was found no multicollinearity (mean value = 203 1.13). Both bi-variable and multi-variable logistic regressions were employed. All explanatory 204 variables in binary logistic regression with a p-value of 0.25 and below were considered candidate 205 variables for multivariable logistic regression analysis to control confounding factors. In the final 206 model, Adjusted Odds Ratio (AOR) with their corresponding 95% confidence intervals (CI) was 207 used to declare factors associated with health system responsiveness. A p-value less than 0.05 was 208 used to declare statistical significance in this study. 209

Socio-demographic characteristics of the study participants 210
A total of 417 outpatients were participated in the study, giving a 98.6% response rate. The median 211 age of the study participants was 33 years with an interquartile range of 25-49 years. And 40.8% 212 were aged between 18-29 years. More than two-thirds (69.6%) of the patients were from rural 213 residency. Of the study participants, 92.8% were Orthodox Christian followers (Table 1). 214 Notes: *=student, private employee, daily laborer **=single, divorced, windowed ***=Ethiopian 217 Birr (currency) 218 Health service accessibility-related characteristics 219 Nearly sixty eight percent of the participants utilized public health care facilities. More than half 220 (56.1%) had traveled one hour and below to reach the health facility (Table 2). 221

Domains of health system responsiveness 228
The domains of confidentiality and dignity were around 72% good performance. On the other 229 hand, Choice was the least (37.2%) score on the good category of performance (Fig. 2). Factors associated with health system responsiveness 234 Binary logistic regression was employed to evaluate the association between different 235 sociodemographic, health facility related, and patient related variables with health system 236 responsiveness. Variables that were found with a p-value < 0.25 in bivariable logistic regression 237 such as age, occupation, educational status, type of facility, out of pocket payment for transport, 238 perceived satisfaction about health care, and perceived quality of health were found to be a 239 candidate for multivariable logistic regression. Model fitness was tested with Hosmer and 240 Lemeshow Goodness of Fit test (p = 0.52). In the final multivariable logistic regression analysis; 241 type of health facility, OOP payment for transport, and patient satisfaction were significantly 242 associated with HSR. 243 Health system responsiveness among private health care facility users were 8.8 times higher when 244 compared with those who utilized public health facility (AOR: 8.8, 95% CI, 4.32-18.25). 245 Participants who had not paid out of pocket for transport to reach health facility were 1.7 times 246 higher health system responsiveness than their counterparts (AOR: 1.7, 95% CI, 1.03 -2.92). The 247 likelihood of health system responsiveness among satisfied patients were nearly 10 times higher 248 when compared with patients having poor satisfaction (AOR: 9.9, 95% CI, 5.11-19.46) ( Table 4) 249

Discussions 250
The study examined the health system responsiveness and associated factors among outpatients of report. Similarly, the result was higher than a study conducted in Shewarobit, Ethiopia (55.3%) 257 [37]. This could be differences in the study participants, in this study we investigated HSR among 258 all outpatients in the District from each primary health care facility in the District, however, in 259 Shewarobit the study was conducted on case-specific responsiveness among HIV positive 260 individuals. Additionally, the observed better responsiveness might be a result of the government's 261 ongoing efforts to improve service delivery. On the contrary, the finding was lower than a study 262 conducted in Brazil (80%) [11]. This is possibly due to the differences in health care availability 263 and accessibility where they are better than sub-Saharan Africa. Sociocultural and economic 264 disparities also the possible likelihood for these differences. Probably also the difference in study 265 population wherein Brazil it was conducted among older adults. 266 The performance of the responsiveness of health care utilization has differed across each domain. 267 Elsewhere studies in German [44], Thailand [24], and India [17] age was significantly associated 292 with health system responsiveness. 293 Health system responsiveness depends on financial aspects [45]. WHO suggested that travel time 294 was a major contributor to poor responsiveness [15]. Supporting to this idea our study showed that 295 the odds of HSR among participants with no out of pocket payment for transport to reach the health 296 facility was 1.7 times higher than its counterparts. This could probably because the rating of HSR 297 might be influenced by the expectations against relative total worth of expense in obtaining needed 298 health care. As improvement in financial fairness health facilities could rate more responsive [28] 299 300 From the finding of this study, it has clearly shown that the likelihood of HSR among participants 301 who were utilized private health facilities were nearly 9 times higher compared to public health 302 facility utilizers. Similar to this, findings from the African countries' of Ghana [46] and South 303 Africa [31] suggested that the overall responsiveness of public health services was lower compared 304 to private services. The possible reason for the highest responsiveness in private facilities might 305 be due to differences in good patient-physician interaction. Thus, private facilities have the aim to 306 maximize their profit to achieve this objective, they are more responsive to attract clients. 307 When clients were dissatisfied with health outcomes, responsiveness mean sum scores will become 308 low [6,47]. In agreement with this idea, this study observed that clients who had good satisfaction 309 with the health care offered had higher HSR in relation to poorly satisfied individuals. Elsewhere 310 studies in Ghana, Ethiopia [28, 29, 37] also indicated that the more satisfaction the higher the 311 responsiveness. Additionally, the world health organization also suggested that except 312 confidentiality all the domains of health system responsiveness were positively and significantly 313 related to satisfaction [48]. Perhaps because as patients satisfied with a non-medical aspects of 314 care, associated with better compliance and understanding of all the interactions of results. To 315 achieve a higher level of welfare on non-health enhancing aspects of care, a greater health system 316 responsiveness is needed. 317

318
There might not be recall bias since the data were collected immediately after they get health care 319 services on their way to home. 320  The data were collected only from the patient perspective or did not include the providers' 321 perspective 322  If the research was performed with a mixed approach, it could be better. 323  We also acknowledge the response bias because of the self-reported data to minimize it short and 324 interval questionnaires were employed. 325  Because of the cross-sectional nature causal relationships between satisfaction, facility type and 326 payment for transport with health system responsiveness cannot be established. 327

Conclusion 328
This study contributes to health system responsiveness research in Ethiopia among outpatients at 329 primary health care facilities. Even though relatively higher health system responsiveness than 330 case specific study in Ethiopia, the result showed that only confidentiality and dignity domains 331 found the highest score. Overall, HSR was higher in private than public healthcare facilities, 332 additionally satisfied clients and those who didn't pay for transport on their way to the health 333 facility were better responsive than their counterparts. The domain of Autonomy, Waiting time, 334 Basic amenities, Choice were identified as failed to meet the legitimate expectation of the clients 335 regarding the non-health aspects of medical care. They need effort to raise responsiveness of health 336 care service in the District. In addition to this, enhancing patient satisfaction, using input from 337 service users, sharing experience and working with collaboration from private clinics and giving 338 attention to distant coming patients will be important interventions to improve HSR. 339  Note: *= out of pocket 511  Key: HC, health center: P1, Private clinic1: p2, private clinic2 523