Key Points
Question
How frequently do women in Ghana bypass the health care facility nearest their homes in favor of more distant facilities perceived to offer better care, and what factors are associated with this choice?
Findings
In this survey study including 1946 reproductive-aged (ie, 16-49 years) women in Ghana in 2017, 32% of women reported bypassing their nearest health care facility. Women who bypassed sought care at hospitals and private facilities more frequently and paid nearly 2-fold in out-of-pocket costs for care compared with women who did not bypass.
Meaning
These findings suggest that the services offered at many primary care facilities in Ghana may not be meeting the needs of women.
This survey study examines the prevalence, individual characteristics and experiences, and costs associated with bypassing the nearest health care facility among women in Ghana.
Abstract
Importance
Recent reports have highlighted that expanding access to health care is ineffective at meeting the goal of universal health coverage if the care offered does not meet a minimum level of quality. Health care facilities nearest to patient’s homes that are perceived to offer inadequate or inappropriate care are frequently bypassed in favor of more distant private or tertiary-level hospital facilities that are perceived to offer higher-quality care.
Objective
To estimate the frequency with which women in Ghana bypass the nearest primary health care facility and describe patient experiences, costs, and other factors associated with this choice.
Design, Setting, and Participants
This nationally representative survey study was conducted in 2017 and included 4203 households to identify women in Ghana aged 15 to 49 years (ie, reproductive age) who sought primary care within the last 6 months. Women who sought care within the past 6 months were included in the study. Data were analyzed from 2018 to 2019.
Exposures
Bypass was defined as a woman’s report that she sought care at a health facility other than the nearest facility.
Main Outcomes and Measures
Sociodemographic characteristics, reasons why women sought care, reasons why women bypassed their nearest facility, ratings for responsiveness of care, patient experience, and out-of-pocket costs. All numbers and percentages were survey-weighted to account for survey design.
Results
A total of 4289 women met initial eligibility criteria, and 4207 women (98.1%) completed the interview. A total of 1993 women reported having sough health care in the past 6 months, and after excluding those who were ineligible and survey weighting, the total sample included 1946 women. Among these, 629 women (32.3%) reported bypassing their nearest facilities for primary care. Women who bypassed their nearest facilities, compared with women who did not, were more likely to visit a private facility (152 women [24.5%] vs 202 women [15.6%]) and borrow money to pay for their care (151 women [24.0%] vs 234 women [17.8%]). After adjusting for covariates, women who bypassed reported paying a mean of 107.2 (95% CI, 79.1-135.4) Ghanaian Cedis (US $18.50 [95% CI, $13.65-$23.36]) for their care, compared with a mean of 58.6 (95% CI, 28.1-89.2) Ghanaian Cedis (US $10.11 [95% CI, $4.85-15.35]) for women who did not bypass (P = .006). Women who bypassed cited clinician competence (136 women [34.3%]) and availability of supplies (93 women [23.4%]) as the most important factors in choosing a health facility.
Conclusions and Relevance
The findings of this survey study suggest that bypassing the nearest health care facility was common among women in Ghana and that available services at lower levels of primary care are not meeting the needs of a large proportion of women. Among the benefits women perceived from bypassing were clinician competence and availability of supplies. These data provide insights to policy makers regarding potential gaps in service delivery and may help to guide primary health care improvement efforts.
Introduction
Recent global initiatives have embraced ambitious new goals to improve primary health care service delivery and achieve universal health coverage. Three 2018 reports1,2,3 have highlighted that significant work remains to be done to measure and improve quality and that merely expanding access to care is insufficient if the care provided does not meet patients’ needs.
Evidence shows that local health care facilities are frequently not used or are bypassed by individuals in favor of secondary or tertiary-level hospital facilities even when care could be managed at the local primary health care level.4,5,6,7,8 This underutilization of dedicated primary care facilities may lead to increased travel and cost burdens for patients, increased costs to health systems through inefficient utilization, overburdening of higher level facilities, and subversion of the goals of universal health coverage.4,9,10,11
The prevalence of bypassing varies widely within and between countries and health systems.2,4,6,8,9,10,11,12,13,14,15,16,17,18,19 Many individuals bypass what they perceive to be understaffed and poorly equipped public government-run health facilities to seek care in private facilities and hospitals.10 Several studies have described substantial variation in individual behavior across different health systems, particularly with regard to childbirth facilities.4,5,6,11,14,17,20,21,22,23,24,25,26,27,28,29,30,31,32,33
There is limited evidence characterizing the extent of primary health care bypass in Ghana and West Africa in general. A better understanding of these bypass behaviors may provide insights into local primary health care improvement efforts for policy makers and facility managers. Recent surveys by Performance Monitoring and Accountability 2020 (PMA2020)34,35 provide nationally-representative data on care seeking and utilization of primary health care in Ghana. Using these data, we explored several questions. First, what is the prevalence of bypassing primary health care facilities in Ghana? Second, how did bypassing primary health care facilities vary by individual demographic and geographic characteristics? Third, what factors were associated with why and where individuals choose to seek care? Fourth, did the self-reported experience of individuals who bypassed differ from those who did not bypass? Finally, how were individuals’ financial costs associated with bypassing?
Methods
This study was approved by School of Medical Sciences, Komfo Anokye Teaching Hospital Committee on Human Research Publications and Ethics, the Partners Human Research Committee, and the Johns Hopkins School of Public Health institutional review board. All study participants provided written informed consent. This study is reported following the American Association for Public Opinion Research (AAPOR) reporting guideline.
Study Area and Data Collection
Extended description of the data collection methods used by PMA2020 can be found elsewhere.34,35 The survey was designed to provide nationally representative data on key health and development indicators in Ghana. The survey was powered to estimate the modern contraceptive prevalence in women of reproductive age (ie, 15-49 years). Data were collected in partnership with the Kwame Nkrumah University of Science and Technology and included a specially designed primary health care module integrated in the existing PMA2020 facility and household surveys used to track progress toward family planning use targets. Surveys were administered in English and local languages. To obtain the data, the Ghana Statistical Service selected 100 enumeration areas across all 10 regions with probability proportional to size using a master sampling frame stratified by urban-rural areas. Within these enumeration areas, 42 households were selected using a random start method to complete the household survey, and all members of the household were surveyed. All data used in this study were collected during round 6 of the PMA2020 survey conducted in late 2017.
Bypass Definition
In this study, bypass was defined as a woman’s report that she sought care at a health facility other than the nearest facility to her residence (“Is this facility the closest facility to your place of residence?” with binary yes or no answer). The woman was then categorized as having bypassed her nearest facility or not.
Variables
We examined 5 categories of variables and their association with bypassing: demographic characteristics (ie, age, marital status, educational attainment, household wealth, residence [urban or rural]), utilization of care (ie, facility type visited, reasons for seeking care, for whom the care was sought, factors considered most important in the choice of health facility), financial costs, responsiveness of care,36 and self-reported patient experience. A responsiveness index was calculated as the mean score from a woman’s rating of all responsiveness questions. The responsiveness index was divided into quintiles and dichotomized into highest quintile vs all others. A detailed list of variables and descriptions for each category can be found in the eAppendix in the Supplement.
Statistical Analysis
All women who responded to the survey and had visited a health facility in the last 6 months were included in the analyses. To generate nationally representative, population-based estimates, we used survey-weighted summary statistics that accounted for the multistage clustered survey design in all analyses.
We used Poisson regression models with a log-link and robust SEs to estimate the associations of bypass (main variable) with women’s self-reported experience of their care. Outcomes were dichotomized as highest or most positive ratings vs all other ratings combined (a top-box categorization). The exponentiated regression coefficients from the Poisson regression can be interpreted as relative risks (RRs) comparing the proportion of those giving the highest rating between women who bypassed and those who did not. We used linear regression with robust SEs to estimate mean out-of-pocket costs paid by women who bypassed and those who did not. All models were first fit unadjusted for other variables, and then with adjustment for demographic and geographic factors and reasons for seeking care (eAppendix in the Supplement). P values were 2-sided and reserved for model-based analyses and not descriptive statistics to avoid issues of multiple testing. Statistical significance was set at P < .05. We performed analyses in Stata version 15.1 (StataCorp) and SAS version 9.1 (SAS Institute) statistical software. Data were analyzed from 2018 to 2019.
Results
A total of 4289 women met eligibility criteria, and 4207 women (98.1%) were available, consented to the interview, and completed the interview. All 4207 women were asked whether they had visited a health care facility within 6 months prior to the survey. Of those who answered the health care facility question, 1993 women (46.7%) answered yes, and thus completed the subsequent primary health care–specific survey questions. All women answered the bypass question. Of women who had sought health care in the past 6 months, 28 (1.4%) were excluded because the sampled household was not their primary residence. Therefore, the sample of interest for this study included 1965 women who had sought care in the last 6 months and lived in the sample household, which was 1946 women after reweighting to account for the survey design (eFigure in the Supplement). Table 1 provides summary statistics for the sample, stratified by bypass status. Using methods accounting for the multistage sampling design and rounded to the nearest integer, 629 women (32.3%) reported bypassing their nearest facility to seek care.
Table 1. Health Facility Utilization Among Women in Ghana.
| Characteristic | Respondents, No. (%)a | ||
|---|---|---|---|
| Did not bypass nearest facility (n = 1317) | Bypassed nearest facility (n = 629) | Total (N = 1946) | |
| Age, y | |||
| 15-24 | 398 (30.2) | 152 (24.2) | 550 (28.3) |
| 25-34 | 501 (38.1) | 236 (37.6) | 738 (37.9) |
| 35-49 | 418 (31.7) | 240 (38.2) | 658 (33.8) |
| Wealth quintile | |||
| Lowest | 289 (21.9) | 114 (18.1) | 403 (20.7) |
| Lower | 259 (19.7) | 112 (17.9) | 371 (19.1) |
| Middle | 259 (19.6) | 132 (20.9) | 390 (20.0) |
| Higher | 291 (22.1) | 133 (21.2) | 424 (21.8) |
| Highest | 220 (16.7) | 138 (21.9) | 358 (18.4) |
| Education | |||
| Never attended | 220 (16.7) | 106 (16.8) | 326 (16.7) |
| Primary | 233 (17.7) | 119 (18.9) | 352 (18.1) |
| Middle junior secondary school | 521 (39.6) | 232 (36.9) | 754 (38.7) |
| Secondary junior secondary school | 237 (18.0) | 109 (17.3) | 345 (17.8) |
| Higher | 106 (8.1) | 63 (10.1) | 170 (8.7) |
| Marital status | |||
| Currently married | 700 (53.2) | 339 (53.8) | 1039 (53.4) |
| Currently living with partner | 211 (16.0) | 104 (16.5) | 315 (16.2) |
| Divorced | 101 (7.7) | 48 (7.6) | 149 (7.6) |
| Widow | 35 (2.6) | 16 (2.6) | 51 (2.6) |
| Never married | 270 (20.5) | 122 (19.4) | 393 (20.2) |
| Residence | |||
| Urban | 648 (49.2) | 324 (51.4) | 971 (49.9) |
| Rural | 670 (50.8) | 305 (48.6) | 975 (50.1) |
| Region | |||
| Ashanti | 256 (19.4) | 144 (22.9) | 400 (20.6) |
| Brong Ahafo | 75 (5.7) | 36 (5.7) | 110 (5.7) |
| Central | 98 (7.4) | 80 (12.7) | 178 (9.1) |
| Eastern | 184 (14.0) | 45 (7.2) | 229 (11.8) |
| Greater Accra | 210 (15.9) | 81 (12.9) | 291 (14.9) |
| Northern | 126 (9.6) | 76 (12.1) | 202 (10.4) |
| Upper East | 97 (7.4) | 36 (5.7) | 133 (6.8) |
| Upper West | 60 (4.5) | 12 (2.0) | 72 (3.7) |
| Volta | 78 (5.9) | 45 (7.1) | 123 (6.3) |
| Western | 134 (10.2) | 74 (11.7) | 207 (10.7) |
| Facility type visited | |||
| Community-based health planning and services | 196 (15.2) | 14 (2.3) | 210 (11.0) |
| Government hospital or polyclinic | 451 (34.9) | 326 (52.6) | 777 (40.6) |
| Government health center | 366 (28.3) | 97 (15.7) | 464 (24.2) |
| Private hospital or clinic | 202 (15.6) | 152 (24.5) | 353 (18.5) |
| Other | 78 (6.0) | 31 (4.9) | 108 (5.7) |
| Owns a bicycle | 358 (27.2) | 143 (22.8) | 501 (25.8) |
| Owns a car | 115 (8.8) | 66 (10.6) | 182 (9.3) |
| Ability to pay for visit | |||
| Very easy | 335 (25.6) | 166 (26.4) | 500 (25.8) |
| Easy | 628 (48.0) | 235 (37.4) | 863 (44.6) |
| Difficult | 231 (17.6) | 143 (22.7) | 373 (19.3) |
| Very difficult | 115 (8.8) | 85 (13.5) | 200 (10.3) |
| Services covered by insurance | 994 (75.5) | 468 (74.4) | 1462 (75.1) |
| Had to borrow money to pay for visit | 234 (17.8) | 151 (24.0) | 385 (19.8) |
| Paid any fees for family planning last 12 mo | 316 (62.7) | 112 (50.3) | 428 (58.9) |
| Confidence that if you became very sick tomorrow, you would be able to receive effective treatment from the health system | |||
| Not at all confident | 42 (3.2) | 14 (2.2) | 56 (2.9) |
| Not very confident | 96 (7.3) | 39 (6.1) | 135 (7.0) |
| Somewhat confident | 260 (19.8) | 112 (17.8) | 372 (19.2) |
| Very confident | 915 (69.7) | 463 (73.8) | 1378 (71.0) |
| Rating of the amount of time spent with clinician | |||
| Poor | 18 (1.4) | 6 (0.9) | 23 (1.2) |
| Fair | 93 (7.1) | 45 (7.1) | 138 (7.1) |
| Good | 619 (47.0) | 238 (37.9) | 857 (44.0) |
| Very good | 397 (30.1) | 211 (33.5) | 608 (312) |
| Excellent | 190 (14.4) | 130 (20.6) | 320 (16.4) |
| Treated by the same clinician each time | |||
| Never | 92 (7.0) | 71 (11.3) | 163 (8.4) |
| Rarely | 507 (38.6) | 259 (41.2) | 766 (39.4) |
| Frequently | 382 (29.1) | 193 (30.6) | 575 (29.6) |
| Always | 332 (25.3) | 106 (16.9) | 438 (22.6) |
| Sought care for self | 882 (67.0) | 465 (74.0) | 1348 (69.2) |
| Sought care for child | 573 (43.5) | 231 (36.7) | 804 (41.3) |
| Sought care for other | 72 (5.5) | 34 (5.4) | 106 (5.4) |
All data are survey weighted. Counts may not sum to the denominator of the column due to the rounding of survey-weighted data.
Demographic Characteristics Among Women Who Bypassed
A total of 305 of 975 rural women (31.3%) bypassed their nearest health facility, compared to 324 of 971 urban women (33.3%) (Table 1). Women in urban areas tended to be wealthier than those in rural areas, with 335 women (34.5%) in the highest wealth quintile compared with 23 women (2.3%) in rural areas (eTable 1 in the Supplement). There were few women in the highest quintile of wealth in rural areas; however, these women had higher levels of bypassing than the wealthiest women in urban areas (eTable 1 in the Supplement).
Women who bypassed their nearest facility were more likely to be seeking care for themselves (as opposed to for a child or other family member or friend) than women who did not bypass (465 women [73.9%] vs 882 women [67.0%]) (Table 2). Women who went to their nearest facility more frequently reported going for vaccinations than women who bypassed (129 women [10.0%] vs 28 women [4.8%]); however, overall reasons for care-seeking were generally similar between women who bypassed and those who did not (Table 2), suggesting that differing clinical needs were not the major driver of bypass. In particular, the proportions of women who sought care owing to a community health worker referral were similar among women who did not bypass and those who did (13 women [1.0%] vs 4 women [0.7%]). Some women may have self-referred to a more distant facility for strategic reasons. To further examine these patterns, we performed a sensitivity analysis in which we examined strategic bypass. This analysis included only women who bypassed their nearest facility because it was closed or did not offer their desired services. After looking only at this selected group of women, we found no significant differences in our primary findings (eTable 2 in the Supplement).
Table 2. Reasons Women Reported for Seeking Care .
| Reasons care was sought | Respondents, No./total (%)a | ||
|---|---|---|---|
| Did not bypass nearest facility | Bypassed nearest facility | Total | |
| Family planning | |||
| No | 1181/1292 (91.4) | 557/587 (95.0) | 1739/1879 (92.5) |
| Yes | 111/1292 (8.6) | 30/587 (5.0) | 140/1879 (7.5) |
| Maternal needs | |||
| No | 1034/1292 (80.0) | 473/587 (80.6) | 1507/1879 (80.2) |
| Yes | 258/1292 (20.0) | 114/587 (19.4) | 372/1879 (19.8) |
| Vaccine | |||
| No | 1163/1292 (90.0) | 559/587 (95.2) | 1722/1879 (91.6) |
| Yes | 129/1292 (10.0) | 28/587 (4.8) | 157/1879 (8.4) |
| Fever | |||
| No | 730/1292 (56.5) | 356/587 (60.6) | 1086/1879 (57.8) |
| Yes | 562/1292 (43.5) | 231/587 (39.4) | 793/1879 (42.2) |
| Sickness | |||
| No | 1015/1292 (78.5) | 447/587 (76.2) | 1462/1879 (77.8) |
| Yes | 277/1292 (21.5) | 140/587 (23.8) | 417/1879 (22.2) |
| Referred by community health worker | |||
| No | 1279/1292 (99.0) | 583/587 (99.3) | 1862/1879 (99.1) |
| Yes | 13/1292 (1.0) | 4/587 (0.7) | 17/1879 (0.9) |
| Snakebite | |||
| No | 1289/1292 (99.8) | 586/587 (99.8) | 1875/1879 (99.8) |
| Yes | 3/1292 (0.2) | 1/587 (0.2) | 4/1879 (0.2) |
| Injury | |||
| No | 1257/1292 (97.3) | 570/587 (97.1) | 1828/1879 (97.3) |
| Yes | 34/1292 (2.7) | 17/587 (2.9) | 51/1879 (2.7) |
| Blood pressure | |||
| No | 1227/1292 (95) | 548/587 (93.4) | 1775/1879 (94.5) |
| Yes | 65/1292 (5.0) | 39/587 (6.6) | 104/1879 (5.5) |
| Diabetes | |||
| No | 1283/1292 (99.3) | 580/587 (98.8) | 1863/1879 (99.1) |
| Yes | 9/1292 (0.7) | 7/587 (1.2) | 16/1879 (0.9) |
| HIV | |||
| No | 1288/1292 (99.7) | 583/587 (99.4) | 1872/1879 (99.6) |
| Yes | 4/1292 (0.3) | 4/587 (0.6) | 7/1879 (0.4) |
| Eye issue | |||
| No | 1275/1292 (98.7) | 573/587 (97.6) | 1848/1879 (98.3) |
| Yes | 17/1292 (1.3) | 14/587 (2.4) | 31/1879 (1.7) |
| Respiratory condition | |||
| No | 1243/1292 (96.2) | 565/587 (96.2) | 1807/1879 (96.2) |
| Yes | 49/1292 (3.8) | 22/587 (3.8) | 72/1879 (3.8) |
| Abdominal issue | |||
| No | 1230/1292 (95.2) | 546/587 (92.9) | 1776/1879 (94.5) |
| Yes | 62/1292 (4.8) | 41/587 (7.1) | 103/1879 (5.5) |
| Concern about new symptom | |||
| No | 1232/1292 (95.3) | 546/587 (93) | 1777/1879 (94.6) |
| Yes | 60/1292 (4.7) | 41/587 (7.0) | 102/1879 (5.4) |
All data are survey weighted. Counts may not sum to the denominator of the column due to the rounding of survey-weighted data.
Rural women who bypassed typically did not seek care at community-based health planning and services facilities (eTable 1 in the Supplement). Instead, women who bypassed frequently sought care at hospitals or polyclinics and private facilities. Wealthy rural women were also much more likely to bypass their nearest facility than their urban counterparts (eTable 1 in the Supplement).
Factors Associated With Where Women Sought Care
Regardless of bypass status, the 3 highest-ranking factors in choice of health care were competency of the clinician (with clinician defined as anyone who provides clinical care) (346 women [27.4%]), well-supplied facility (273 women [21.7%]), and respect from the clinician (156 women [12.4%]) (Table 3). Among women who bypassed, 136 women (34.3%) reported competence of the clinician at that facility as being the most important factor influencing their decision, compared with 210 women (24.3%) who did not bypass. The distance of the facility from the respondent’s home was another key difference between those who bypassed and those who did not, with just 8 women (2.1%) who bypassed listing it as the most important factor in choosing a care facility, compared with 91 women (10.5%) who did not bypass. Women who bypassed reported a variety of reasons why they bypassed their closest facility, of which the most commonly reported reason was that their closest facility did not provide the services that they needed (234 of 479 respondents [48.9%]) (eTable 2 in the Supplement).
Table 3. Most Important Factors Reported in Choice of Health Facility .
| Factor | Respondents, No. (%)a | ||
|---|---|---|---|
| Did not bypass nearest facility (n = 1309) | Bypassed nearest facility (n = 621) | Total (N = 1930) | |
| Most important overall | |||
| Competency of clinician | 210 (24.3) | 136 (34.3) | 346 (27.4) |
| Facility well supplied | 181 (20.9) | 93 (23.4) | 273 (21.7) |
| Respect from clinician | 109 (12.6) | 47 (11.9) | 156 (12.4) |
| Cost of treatment | 104 (12.0) | 34 (8.6) | 138 (10.9) |
| Waiting time | 71 (8.2) | 30 (7.5) | 101 (8.0) |
| Privacy | 44 (5.1) | 23 (5.9) | 68 (5.4) |
| Cleanliness of facility | 23 (2.7) | 12 (3.1) | 36 (2.8) |
| Distance | 91 (10.5) | 8 (2.1) | 100 (7.9) |
| Choice of clinician | 5 (0.6) | 7 (1.6) | 11 (0.9) |
| Knowledge of clinician | 6 (0.7) | 4 (0.9) | 10 (0.8) |
| Other | 1 (0.1) | 2 (0.4) | 3 (0.2) |
| Cost of visit | 17 (2.0) | 1 (0.3) | 18 (1.5) |
| Traditional medicine practiced | 2 (0.2) | 0 | 2 (0.1) |
| Total | 867 (100) | 396 (100) | 1263 (100) |
| Described as important for choosing facility | |||
| Wait time | |||
| No | 946 (72.3) | 470 (75.6) | 1415 (73.3) |
| Yes | 363 (27.7) | 152 (24.4) | 515 (26.7) |
| Distance | |||
| No | 939 (71.8) | 561 (90.4) | 1501 (77.8) |
| Yes | 369 (28.2) | 60 (9.6) | 429 (22.2) |
| Cleanliness | |||
| No | 931 (71.1) | 456 (73.3) | 1386 (71.8) |
| Yes | 378 (28.9) | 166 (26.7) | 544 (28.2) |
| Respect from clinician | |||
| No | 840 (64.2) | 394 (63.4) | 1234 (63.9) |
| Yes | 469 (35.8) | 228 (36.6) | 696 (36.1) |
| Competence of clinician | |||
| No | 776 (59.3) | 269 (43.3) | 1045 (54.2) |
| Yes | 532 (40.7) | 352 (56.7) | 885 (45.8) |
| Privacy | |||
| No | 1103 (84.3) | 527 (84.9) | 1631 (84.5) |
| Yes | 206 (15.7) | 94 (15.1) | 299 (15.5) |
| Well-supplied facility | |||
| No | 872 (66.6) | 392 (63.0) | 1263 (65.5) |
| Yes | 437 (33.4) | 230 (37.0) | 667 (34.5) |
| Cost of treatment | |||
| No | 925 (70.7) | 461 (74.1) | 1386 (71.8) |
| Yes | 383 (29.3) | 161 (25.9) | 544 (28.2) |
| Cost of visit | |||
| No | 1223 (93.5) | 607 (97.6) | 1830 (94.8) |
| Yes | 86 (6.5) | 15 (2.4) | 100 (5.2) |
| Choice of clinician | |||
| No | 1269 (96.9) | 593 (95.5) | 1862 (96.5) |
| Yes | 40 (3.1) | 28 (4.5) | 68 (3.5) |
| Knowledge of clinician | |||
| No | 1277 (97.5) | 607 (97.7) | 1884 (97.6) |
| Yes | 32 (2.5) | 14 (2.3) | 46 (2.4) |
| Traditional medicine practiced | |||
| No | 1305 (99.7) | 618 (99.4) | 1922 (99.6) |
| Yes | 4 (0.3) | 4 (0.6) | 8 (0.4) |
All data are survey weighted. Counts may not sum to the denominator of the column due to the rounding of survey-weighted data.
Self-reported Experience of Women Who Bypassed vs Those Who Did Not
Women who bypassed generally reported receiving more responsive care (Table 4). In adjusted Poisson regression models, women who bypassed were more likely to rate the following characteristics as excellent compared with nonexcellent ratings: clinician choice (RR, 1.47 [95% CI, 1.13-1.90]; P = .004) and cleanliness (RR, 1.36 [95% CI, 1.08-1.71]; P = .01). After covariate adjustment, women who bypassed were less likely to report excellent physical health (RR, 0.87 [95% CI, 0.76-1.00]; P = .04). The full list of adjustment variables is presented in Table 4.
Table 4. Highest Ratings of Responsiveness, PROMs, and Affordability of Care Comparing Women Who Bypassed Their Nearest Health Care Facility With Those Who Did Not Using Poisson Regression With a Log-Link and Robust SEs.
| Rating | Unadjusted model | Adjusted modela | ||
|---|---|---|---|---|
| Relative risk (95% CI) | P value | Relative risk (95% CI) | P value | |
| Responsiveness | ||||
| Wait rating | 1.08 (0.72-1.62) | .71 | 0.99 (0.67-1.47) | .98 |
| Respect from clinician | 1.41 (0.97-2.06) | .07 | 1.14 (0.83-1.55) | .41 |
| Involvement | 1.70 (1.03-2.79) | .04 | 1.23 (0.77-1.98) | .38 |
| Provider choice | 1.64 (1.08-2.50) | .02 | 1.47 (1.13-1.90) | .004 |
| Cleanliness | 1.76 (1.26-2.47) | .001 | 1.36 (1.08-1.71) | .01 |
| Understand clinician | 1.61 (1.06-2.43) | .02 | 1.24 (0.86-1.77) | .25 |
| Privacy | 1.73 (1.12-2.68) | .01 | 1.37 (1.00-1.87) | .05 |
| Responsiveness index | 1.56 (1.03-2.36) | .04 | 1.25 (0.92-1.70) | .14 |
| PROM | ||||
| Excellent quality of care | 1.21 (1.04-1.41) | .01 | 1.04 (0.94-1.17) | .44 |
| Overall excellent quality of care received | 1.14 (0.98-1.34) | .10 | 0.99 (0.88-1.10) | .80 |
| Physical health rating | 0.95 (0.80-1.14) | .57 | 0.87 (0.76-1.00) | .04 |
| Mental health rating | 1.06 (0.83-1.34) | .64 | 0.93 (0.76-1.13) | .45 |
| Recommend this facility to others | 1.01 (0.99-1.03) | .37 | 1.01 (0.99-1.03) | .38 |
| Affordability of care | ||||
| Had to borrow money or sell things to pay for visit | 1.35 (1.04-1.74) | .02 | 1.36 (1.10-1.68) | .005 |
Abbreviation: PROM, patient-reported outcome measure.
Adjusted for age, wealth quintile, highest level of education attained, urban or rural neighborhood, confidence that if you became very sick tomorrow that you would be able to receive effective treatment from the health system, taken care of by the same clinician each visit, reason for seeking care, person for whom care was sought, facility type, and region.
Costs Associated With Bypassing
Women who bypassed were more likely to have borrowed money to pay for their care (RR, 1.36 [95% CI, 1.10-1.68]; P = .005) (Table 4). Using linear regression and adjusting for reasons for seeking care and other factors, women who bypassed reported paying a mean of 107.2 (95% CI, 79.1-135.4) Ghanaian Cedis (US $18.50 [95% CI, $13.65-$23.36]) for their care, compared with a mean of 58.6 (95% CI, 28.1-89.2) Ghanaian Cedis (US $10.11 [95% CI, $4.85-15.35]) for women who did not bypass (P = .006) (Table 5). Even after adjustment for other factors, including reason for seeking care, women who bypassed paid nearly 2-fold as much for their care as those who did not.
Table 5. Mean Costs of Care Using Generalized Linear Models With Robust SEs.
| Bypassed nearest facility | Unadjusted model | Adjusted modela | ||
|---|---|---|---|---|
| Cost, GH₵ (95% CI) | P value | Cost, GH₵ (95% CI) | P value | |
| No | 53.1 (29.3-76.3) | <.001 | 58.6 (28.1-89.2) | .006 |
| Yes | 117.0 (74.1-159.9) | 107.2 (79.1-135.4) | ||
Abbreviation: GH₵, Ghanaian Cedi.
Adjusted for age, wealth quintile, highest level of education attained, urban or rural neighborhood, confidence that if you became very sick tomorrow, you would be able to receive effective treatment from the health system, cared for by the same clinician each visit, reason for seeking care, person for whom care was sought, facility type, and region.
Discussion
This survey study used a nationally representative survey to estimate the countrywide prevalence of health facility bypass in Ghana. Unlike previous studies that focused on childbirth or emergency services, our analysis focused on a general population’s experience of primary care and offers insights into who bypasses, why, and the outcomes associated with their care. We found that bypassing was relatively common: 32.3% of women reported bypassing. Moreover, bypassing was mostly owing to the knowledge that the desired services were not available at the closest facility. Women who bypassed cited clinician competence as the most important factor in choosing a facility more frequently than women who did not bypass, indicating that women highly value competent clinicians and are willing to travel farther to get them. In rural areas, wealthier women tended to bypass more frequently, suggesting differences in behavior and functional access between socioeconomic strata.
Women who bypassed paid more out of pocket for their care than women who did not bypass. To our knowledge, this is the first study to examine costs associated with bypass in this context. Other studies have examined discrete choice experiments of health facility utilization and quality and found that patients are generally willing to pay more for higher-quality services, although how much more is greatly dependent on context.20,21 Kahabuka et al7 found that caretakers in Tanzania frequently bypassed nearest facilities for their children owing to lack of diagnostic equipment, insufficient drug stocks, limited hours, and poor services—similar to our findings. However, comparisons of bypass across countries can vary widely depending on the service being sought and the distribution of facilities offering that service. For example, in a 2020 study of family planning in Tanzania,37 67% of women bypassed their nearest facility to seek contraception. The nearest facility was a mean of 1.2 km away, while the chosen facility was a mean of 2.9 km away. In contrast, a 2016 study of utilization of childbirth facilities in Ghana38 found that women lived a median 3.3 km from their nearest birth facility, with the nearest hospital 13.9 km away (bypass was not assessed). These differences reflect heterogeneities in opportunity for choice of facility, proximity of facilities offering a desired service, and burden associated with bypassing the nearest facility. Our population of women was also fairly young and as such, less likely to be seeking care for the types of chronic conditions and serious illnesses that might be associated with referral to hospitals.
Women generally reported similar reasons for seeking care whether they bypassed their nearest facility or not. This may suggest that many women’s reasons for bypassing were less associated with the nature of their underlying health concerns and more associated with their perception of quality of care or experiences at the facility. There are limited published data from other low- and middle-income countries on differences between individuals who bypass and those who do not in reasons why individuals might seek care. A 2011 study by Gauthier et al11 examined reasons for seeking care by facility type in Chad and reported that malaria and diarrhea were the primary reasons for seeking care; however, the study did not assess whether these reasons differed by bypass status.
Previous studies4,6,9,10,11,18,19,39 have suggested that bypass is frequently driven by perceptions of the quality of care offered. In our study, women who bypassed were more likely to rate their clinician choice, cleanliness of the facility, and privacy as excellent. However, despite rating these aspects of their facility highly, women who bypassed were no more likely to report excellent ratings for the overall quality of their care or to recommend the facility highly to others. This finding may be associated with differential expectations of what is considered to be acceptable care, with individuals who bypass having higher expectations than whose who do not.40 In previous studies,18,41,42 vignettes describing aspects of care indicated that individuals have wide variation in what they consider to be acceptable care, particularly those who are poorer and less educated.
Taken in the context of a growing emphasis on quality in health care delivery, our results have important policy and public health implications. They suggest that bypassing is common, expensive, and frequently associated with perceptions that the closest facilities do not provide the desired services. Women frequently reported that the competence of their clinician was of highest importance, and were most likely to bypass community-based health planning and services facilities, which do not have physicians on staff. Increased staff training, higher quality of care, and range and readiness of available services may decrease bypassing, reducing costs to patients.7 Further research is necessary to identify which specific services can be brought closer to patients, if possible. Our results also indicate an opportunity to improve the system’s efficiency in being a point of first contact for patients, per Starfield’s four pillars of primary care43 (ie, first contact, continuous care, comprehensive care, and coordinated care). While further research in this area is essential, improving the quality of care at local levels may help toward building more equitable, higher quality primary health care systems that better address patients’ needs.44,45
Limitations
Our study has limitations. First, this is a cross-sectional survey designed to be nationally representative of reproductive-aged women in Ghana, so our findings should be interpreted with caution and may not be generalizable to other subpopulations. Second, our definition of bypass was based on women’s self-report of not visiting their nearest facility, which may be inaccurate, as many women may have been unaware of precise distances between their home and facilities. However, this definition has been used in previous studies, and our assumption is that women’s perception of distance, rather than actual distance, is likely to be more relevant to their choice of health care facility.6,9,19,30 Third, although our study focused on women who visited a facility, some women reported that they were unable to visit a health facility altogether, despite a desire for care. Fourth, because our study collected only limited data on reasons why women may have bypassed their closest facility, care should be taken in drawing definitive policy and public health conclusions based on our findings alone.
Conclusions
In this survey study of health care utilization among reproductive-aged women in Ghana, we found that bypass was relatively common in Ghana, with nearly one-third of women bypassing the facility closest to their home. Women who bypassed their nearest health facility cited the competence of the clinician most frequently as the most important factor in facility choice. Women who bypassed were also more likely to visit hospitals and private clinics than women who did not bypass. Women who bypassed paid nearly 2-fold as much for their care as women who did not bypass. These data provide insights that may be helpful for policy makers and facility managers in Ghana as they strive to achieve high-quality universal health care for the population.
eAppendix. Detailed Description of Variables
eFigure. Extent of Bypassing of Health Care Facilities in Ghana in 2017
eTable 1. Ghanaian Health Facility Utilization by Urban and Rural Neighborhood
eTable 2. Reasons Given for Why Women Bypassed Their Nearest Facility
References
- 1.National Academies of Sciences, Engineering, and Medicine Crossing the Global Quality Chasm: Improving Health Care Worldwide. National Academies Press; 2018. doi: 10.17226/25152 [DOI] [PubMed] [Google Scholar]
- 2.Organisation for Economic Co-operation and Development, World Health Organization, World Bank Group . Delivering Quality Health Services: A Global Imperative. World Health Organization; 2018. doi: 10.1787/9789264300309-en [DOI] [Google Scholar]
- 3.Kruk ME, Gage AD, Arsenault C, et al. . High-quality health systems in the sustainable development goals era: time for a revolution. Lancet Glob Health. 2018;6(11):e1196-e1252. doi: 10.1016/S2214-109X(18)30386-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Leonard KL, Mliga GR, Haile Mariam D. Bypassing health centres in Tanzania: revealed preferences for quality. J Afr Econ. 2002;11(4):441-471. doi: 10.1093/jae/11.4.441 [DOI] [Google Scholar]
- 5.Aoki T, Yamamoto Y, Ikenoue T, et al. . Effect of patient experience on bypassing a primary care gatekeeper: a multicenter prospective cohort study in Japan. J Gen Intern Med. 2018;33(5):722-728. doi: 10.1007/s11606-017-4245-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Rao KD, Sheffel A. Quality of clinical care and bypassing of primary health centers in India. Soc Sci Med. 2018;207:80-88. doi: 10.1016/j.socscimed.2018.04.040 [DOI] [PubMed] [Google Scholar]
- 7.Kahabuka C, Kvåle G, Moland KM, Hinderaker SG. Why caretakers bypass primary health care facilities for child care: a case from rural Tanzania. BMC Health Serv Res. 2011;11:315. doi: 10.1186/1472-6963-11-315 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Liu JJ, Bellamy G, Barnet B, Weng S. Bypass of local primary care in rural counties: effect of patient and community characteristics. Ann Fam Med. 2008;6(2):124-130. doi: 10.1370/afm.794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Akin JS, Hutchinson P. Health-care facility choice and the phenomenon of bypassing. Health Policy Plan. 1999;14(2):135-151. doi: 10.1093/heapol/14.2.135 [DOI] [PubMed] [Google Scholar]
- 10.Filmer D, Hammer JS, Pritchett LH. Weak links in the chain: a diagnosis of health policy in poor countries. World Bank Res Obs. 2000;15(2):199-224. doi: 10.1093/wbro/15.2.199 [DOI] [Google Scholar]
- 11.Gauthier B, Wane W. Bypassing health providers: the quest for better price and quality of health care in Chad. Soc Sci Med. 2011;73(4):540-549. doi: 10.1016/j.socscimed.2011.06.008 [DOI] [PubMed] [Google Scholar]
- 12.Kruk ME, Hermosilla S, Larson E, Mbaruku GM. Bypassing primary care clinics for childbirth: a cross-sectional study in the Pwani region, United Republic of Tanzania. Bull World Health Organ. 2014;92(4):246-253. doi: 10.2471/BLT.13.126417 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shah R. Bypassing birthing centres for child birth: a community-based study in rural Chitwan Nepal. BMC Health Serv Res. 2016;16(1):597. doi: 10.1186/s12913-016-1848-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kanté AM, Exavery A, Phillips JF, Jackson EF. Why women bypass front-line health facility services in pursuit of obstetric care provided elsewhere: a case study in three rural districts of Tanzania. Trop Med Int Health. 2016;21(4):504-514. doi: 10.1111/tmi.12672 [DOI] [PubMed] [Google Scholar]
- 15.Sabde Y, Chaturvedi S, Randive B, et al. . Bypassing health facilities for childbirth in the context of the JSY cash transfer program to promote institutional birth: a cross-sectional study from Madhya Pradesh, India. PLoS One. 2018;13(1):e0189364. doi: 10.1371/journal.pone.0189364 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kruk ME, Mbaruku G, McCord CW, Moran M, Rockers PC, Galea S. Bypassing primary care facilities for childbirth: a population-based study in rural Tanzania. Health Policy Plan. 2009;24(4):279-288. doi: 10.1093/heapol/czp011 [DOI] [PubMed] [Google Scholar]
- 17.Yaffee AQ, Whiteside LK, Oteng RA, et al. . Bypassing proximal health care facilities for acute care: a survey of patients in a Ghanaian accident and emergency centre. Trop Med Int Health. 2012;17(6):775-781. doi: 10.1111/j.1365-3156.2012.02984.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kruk ME, Gage AD, Joseph NT, Danaei G, García-Saisó S, Salomon JA. Mortality due to low-quality health systems in the universal health coverage era: a systematic analysis of amenable deaths in 137 countries. Lancet. 2018;392(10160):2203-2212. doi: 10.1016/S0140-6736(18)31668-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kujawski SA, Leslie HH, Prabhakaran D, Singh K, Kruk ME. Reasons for low utilisation of public facilities among households with hypertension: analysis of a population-based survey in India. BMJ Glob Health. 2018;3(6):e001002. doi: 10.1136/bmjgh-2018-001002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hanson K, McPake B, Nakamba P, Archard L. Preferences for hospital quality in Zambia: results from a discrete choice experiment. Health Econ. 2005;14(7):687-701. doi: 10.1002/hec.959 [DOI] [PubMed] [Google Scholar]
- 21.Larson E, Vail D, Mbaruku GM, Kimweri A, Freedman LP, Kruk ME. Moving toward patient-centered care in Africa: a discrete choice experiment of preferences for delivery care among 3,003 Tanzanian women. PLoS One. 2015;10(8):e0135621. doi: 10.1371/journal.pone.0135621 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fleming LC, Ansumana R, Bockarie A, et al. . Inpatient healthcare provider bypassing by women and their children in urban Bo, Sierra Leone. Pan Afr Med J. 2016;23:146. doi: 10.11604/pamj.2016.23.146.8706 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Salazar M, Vora K, Costa AD. Bypassing health facilities for childbirth: a multilevel study in three districts of Gujarat, India. Glob Health Action. 2016;9:32178. doi: 10.3402/gha.v9.32178 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Okoli H, Obembe T, Osungbade K, Adeniji F, Adewole D. Self-referral patterns among federal civil servants in Oyo state, South-Western Nigeria. Pan Afr Med J. 2017;26:105. doi: 10.11604/pamj.2017.26.105.11483 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sun X, Meng H, Ye Z, Conner KO, Duan Z, Liu D. Factors associated with the choice of primary care facilities for initial treatment among rural and urban residents in Southwestern China. PLoS One. 2019;14(2):e0211984. doi: 10.1371/journal.pone.0211984 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Koce F, Randhawa G, Ochieng B. Understanding healthcare self-referral in Nigeria from the service users’ perspective: a qualitative study of Niger state. BMC Health Serv Res. 2019;19(1):209. doi: 10.1186/s12913-019-4046-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yao J, Agadjanian V. Bypassing health facilities in rural Mozambique: spatial, institutional, and individual determinants. BMC Health Serv Res. 2018;18(1):1006. doi: 10.1186/s12913-018-3834-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Escamilla V, Calhoun L, Winston J, Speizer IS. The role of distance and quality on facility selection for maternal and child health services in urban Kenya. J Urban Health. 2018;95(1):1-12. doi: 10.1007/s11524-017-0212-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Montagu D, Sudhinaraset M, Diamond-Smith N, et al. . Where women go to deliver: understanding the changing landscape of childbirth in Africa and Asia. Health Policy Plan. 2017;32(8):1146-1152. doi: 10.1093/heapol/czx060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Karkee R, Lee AH, Binns CW. Bypassing birth centres for childbirth: an analysis of data from a community-based prospective cohort study in Nepal. Health Policy Plan. 2015;30(1):1-7. doi: 10.1093/heapol/czt090 [DOI] [PubMed] [Google Scholar]
- 31.Gage AD, Carnes F, Blossom J, et al. . In low- and middle-income countries, is delivery in high-quality obstetric facilities geographically feasible? Health Aff (Millwood). 2019;38(9):1576-1584. doi: 10.1377/hlthaff.2018.05397 [DOI] [PubMed] [Google Scholar]
- 32.Gage AD, Leslie HH, Bitton A, et al. . Does quality influence utilization of primary health care: evidence from Haiti. Global Health. 2018;14(1):59. doi: 10.1186/s12992-018-0379-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cohen J, Golub G, Kruk ME, McConnell M. Do active patients seek higher quality prenatal care: a panel data analysis from Nairobi, Kenya. Prev Med. 2016;92:74-81. doi: 10.1016/j.ypmed.2016.09.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Zimmerman L, Olson H, Tsui A, Radloff S; PMA2020 Principal Investigators Group . PMA2020: rapid turn-around survey data to monitor family planning service and practice in ten countries. Stud Fam Plann. 2017;48(3):293-303. doi: 10.1111/sifp.12031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zimmerman LA, Bell SO, Li Q, Morzenti A, Anglewicz P, Tsui AO; PMA2020 Principal Investigators Group . Individual, community and service environment factors associated with modern contraceptive use in five Sub-Saharan African countries: a multilevel, multinomial analysis using geographically linked data from PMA2020. PLoS One. 2019;14(6):e0218157. doi: 10.1371/journal.pone.0218157 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Darby C, Valentine N, De Silva A, Murray CJL; World Health Organization World Health Organization (WHO): strategy on measuring responsiveness. Accessed July 8, 2020. https://www.who.int/responsiveness/papers/paper23.pdf
- 37.Elewonibi B, Sato R, Manongi R, Msuya S, Shah I, Canning D. The distance-quality trade-off in women’s choice of family planning provider in North Eastern Tanzania. BMJ Glob Health. 2020;5(2):e002149. doi: 10.1136/bmjgh-2019-002149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nesbitt RC, Lohela TJ, Soremekun S, et al. . The influence of distance and quality of care on place of delivery in rural Ghana. Sci Rep. 2016;6:30291. doi: 10.1038/srep30291 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Liu L, Leslie HH, Joshua M, Kruk ME. Exploring the association between sick child healthcare utilisation and health facility quality in Malawi: a cross-sectional study. BMJ Open. 2019;9(7):e029631. doi: 10.1136/bmjopen-2019-029631 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Awoke MA, Negin J, Moller J, et al. . Predictors of public and private healthcare utilization and associated health system responsiveness among older adults in Ghana. Glob Health Action. 2017;10(1):1301723. doi: 10.1080/16549716.2017.1301723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Macarayan EK, Gage AD, Doubova SV, et al. . Assessment of quality of primary care with facility surveys: a descriptive analysis in ten low-income and middle-income countries. Lancet Glob Health. 2018;6(11):e1176-e1185. doi: 10.1016/S2214-109X(18)30440-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Roder-DeWan S, Gage AD, Hirschhorn LR, et al. . Expectations of healthcare quality: a cross-sectional study of internet users in 12 low- and middle-income countries. PLoS Med. 2019;16(8):e1002879. doi: 10.1371/journal.pmed.1002879 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Starfield B. Primary Care: Concept, Evaluation, and Policy. Oxford University Press; 1992. [Google Scholar]
- 44.Macarayan EK, Ratcliffe HL, Otupiri E, et al. . Facility management associated with improved primary health care outcomes in Ghana. PLoS One. 2019;14(7):e0218662. doi: 10.1371/journal.pone.0218662 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Bitton A, Ratcliffe HL, Veillard JH, et al. . Primary health care as a foundation for strengthening health systems in low- and middle-income countries. J Gen Intern Med. 2017;32(5):566-571. doi: 10.1007/s11606-016-3898-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eAppendix. Detailed Description of Variables
eFigure. Extent of Bypassing of Health Care Facilities in Ghana in 2017
eTable 1. Ghanaian Health Facility Utilization by Urban and Rural Neighborhood
eTable 2. Reasons Given for Why Women Bypassed Their Nearest Facility