Abstract
Purpose:
Access to eye surgical care in low- and middle-income countries (LMICs) remains limited due to geographical and financial barriers. This survey evaluated the travel and financial burden on patients and caregivers attending perioperative cataract care at an urban base hospital (UBH) versus community clinics (vision centers [VC]) at the Aravind Eye Care System in South India.
Methods:
This cross-sectional study surveyed 105 cataract surgery patients divided into three groups based on perioperative appointment location: UBH-only (appointments at UBH), VC-only (appointments at VCs), and UBH/VC (Day 1 postoperative appointment at VCs, others at UBH). Descriptive statistics and linear regression assessed associations between subgroups and travel and financial burden. The UBH/VC group reported their preferred location and the reasons.
Results:
Over the entire appointment period, transport time for VC-only (353 ± 118 min) was over 3 h lower than UBH-only (589 ± 418 min) and UBH/VC (568 ± 230 min; P < 0.001). Total appointment time was lowest for VC-only (562 ± 177; 1069 ± 439 in UBH-only; 1021 ± 383 min in UBH/VC; P < 0.001). Compared to UBH-only, the VC-only group had the lowest transport time (−236 min, 95% CI: −371 to −102, P = 0.001) and total appointment time (−507 min, 95% CI: −673 to −340, P < 0.001). Transport costs and missed wages were lower for VC-only participants for preoperative and postoperative Day 1 appointments (P < 0.001). Among UBH/VC, 63% (n = 22) preferred VC, while 37% (n = 13) preferred UBH.
Conclusions:
Decentralized perioperative follow-up care is associated with reduced travel and financial burdens for cataract surgery patients in rural, low-resource settings. Further research is needed to evaluate the clinical effectiveness and operational feasibility of decentralized postoperative care in LMICs.
Keywords: Cataract, decentralized care, eye health equity, global eye health, healthcare access, surgical care, vision centers
Cataract, the leading cause of global blindness, affects over 35 million people and imposes significant quality-of-life and financial burdens on patients and caretakers.[1] In 2019, cataract led to a loss of 6.68 million disability-adjusted life years (DALYs), making it the eye condition with the greatest disease burden.[2] The aging global population is driving up surgical demand.[3]
The United Nations has set a goal of achieving equitable eye care access by 2030.[4,5] However, cataract blindness disproportionately affects vulnerable populations, with over 90% of DALYs in developing countries, especially in rural areas lacking access to care.[6] Therefore, strategies addressing the availability, affordability, accessibility, and acceptability of eye care are crucial, particularly for low-resource populations.[7] India, a low- to middle-income country (LMIC) with the highest burden of cataract blindness,[8,9] offers a context for studying challenges in accessing eye care.
The Aravind Eye Care System (AECS) in India is one of the largest eye care systems globally, having performed over 7.8 million surgeries across 14 hospitals. The AECS care delivery model prioritizes providing equitable, high-quality care and has been adapted throughout India and other LMICs.[10] AECS provides primary eye care through a network of vision centers (VCs), which are ophthalmic technician-run primary eye care facilities located in rural areas to allow underserved patients to be treated on-site, closer to their homes, without having to travel to urban hospitals several hours away. Technicians perform vision evaluation, refraction, slit-lamp biomicroscopy, applanation tonometry, and fundus evaluation. They can consult ophthalmologists at the urban base hospital (UBH) through live teleconferencing.[11]
Cataract surgery involves multiple appointments: a preoperative surgical planning visit with the surgeon and postoperative follow-ups, usually at 1 day (POD1), 1 week (POW1), and 1 month (POM1) after the surgery. This can create travel burdens for rural patients who must attend their perioperative appointments in urban hospitals.
During the COVID-19 pandemic, AECS piloted a decentralized perioperative care model to avoid patient crowding in urban hospitals, allowing uncomplicated cataract patients to receive perioperative care at multiple VCs located in semi-rural and rural communities across South India instead of UBH.[12] By 2021, approximately half of POD1 cataract patients were followed up at VCs, with less than 10% requiring referral to the UBH for further evaluation.[10]
The nearly equal distribution of follow-up appointments between the UBH and VCs presented an opportunity to compare centralized and decentralized models of cataract perioperative care in terms of travel and economic burden for patients and caretakers. This study quantified the travel and financial burdens associated with cataract appointments, assessing the extent to which decentralized perioperative follow-up at VCs could reduce them for low-resource populations.
Methods
The survey was offered to patients aged ≥18 years who underwent cataract surgery at the Aravind Eye Hospital, Pondicherry, India, referred to as the UBH, in January 2023. Patients who wished to participate provided written informed consent. Recruitment occurred consecutively until a sample size of 35 participants per group was reached. This study adhered to the Declaration of Helsinki, and the protocol received ethics approval from the Aravind Eye Hospital Institutional Review Board.
Participants were divided into three subgroups based on the locations of their perioperative appointments, all of which were single-day visits:
UBH-only: preoperative, POD1, and POM1 appointments at the UBH.
UBH/VC: preoperative and POM1 appointments at the UBH, but POD1 appointments at a VC.
VC-only: preoperative, POD1, and POM1 appointments at VCs.
All participants underwent outpatient cataract surgery, were discharged the same day, and returned for a POD1 visit. Follow-up location was based on personal preference unless complications occurred, in which case follow-up was completed at the UBH. A Tamil-speaking study coordinator conducted a de-identified postoperative survey to collect demographic data, including age, sex, home address with zip code, daily wage, travel accompaniment, and the type of transport (public or private) used to attend the appointment. A Google Maps service was used to input zip code data and estimate the round-trip distance between the UBH and participants’ homes, representing the distance they would have had to travel if they attended their appointments at UBH. For participants with multiple caretakers, information was gathered solely from the primary caretaker responsible for transportation and expenses.
Survey metrics assessed travel burden (transport and total appointment time) and financial burden (missed wage of the patient combined with transport cost). “Transport time” refers to the total time spent exclusively on round-trip travel from home to the UBH or VCs. “Total appointment time” encompasses the transport time as well as all intervening time, such as time spent waiting in health care facilities and time spent during medical appointments at the UBH or VCs. Transport cost encompassed the reported expenses incurred by patients and caregivers for the round-trip between the hospital and home. Missed wages reflected income lost on the day of the follow-up visit only, as all patients were managed as outpatients. UBH/VC patients, who attended both the VC and UBH, were asked about their preferred location and the reason for their choice.
Responses were analyzed using Stata/BE version 17 (StataCorp LLC, College Station, TX, USA). Chi-square tests and analysis of variance were used to compare categorical and continuous variables, respectively. As the normality assumptions for parametric tests were not met for variables such as distance travelled for a round trip from home to the UBH and wage, the Fisher exact test and Kruskal–Wallis test were used to assess group differences. ß coefficients and 95% confidence intervals (CI) from univariate regression were calculated to examine the association of each subgroup (UBH-only, UBH/VC, VC-only) with travel burden (transport time and total appointment time), using UBH-only as the reference group. A P value <0.05 was considered statistically significant.
Results
Patient demographics
We surveyed 105 participants in January 2023 (35 participants each in UBH-only, UBH/VC, and VC-only). Table 1 shows demographic characteristics. The mean age of participants was 61 ± 12 years (IQR 55–70) in UBH-only, 61 ± 9 years (IQR 54–66) in UBH/VC, and 61 ± 9 years (IQR 55–67) in VH-only groups, with no significant differences between the groups (P = 0.94). Sex distribution was similar, with 46% (n = 16) males in UBH-only, 54% (n = 19) males in UBH/VC, and 40% (n = 14) males in VC-only (P = 0.52). VC-only participants lived further away from UBH, with the estimated round-trip distance from participants’ homes to the UBH being 92 ± 94 km for UBH-only, 147 ± 72 km for UBH/VC, and 153 ± 76 km for VC-only (P = 0.25). UBH/VC had the highest mean wage of 812 ± 1063 INR (UH-only: 709 ± 1331; VC-only: 202 ± 331 INR; P < 0.01). The proportion of participants with an accompanying person during their appointments and the proportion using public versus private transport were similar across groups for all visits (P > 0.05).
Table 1.
Baseline characteristics of study population
| UBH-only (n=35) | UBH/VC (n=35) | VC-only (n=35) | P a | |
|---|---|---|---|---|
| Age, mean (SD) | 61 (12) | 61 (9) | 61 (9) | 0.94 |
| Sex, n (%) | ||||
| Male | 16 (46) | 19 (54) | 14 (40) | 0.52 |
| Female | 19 (54) | 16 (46) | 21 (60) | |
| Estimated round-trip distance from participants’ homes to UBH, mean (SD)b | 92 (94) | 147 (72) | 153 (76) | 0.25 |
| Wage, mean (SD)c | 709 (1331) | 812 (1063) | 202 (331) | <0.01 |
| Adjusted wage, mean (SD)c,d | 561 (388) | 1115 (1167) | 294 (296) | <0.01 |
| Participant occupation category, n (%) | ||||
| Retired/Pensioner | 11 (31) | 7 (20) | 2 (6) | 0.14 |
| No occupation | 8 (23) | 13 (37) | 16 (46) | |
| Skilled farm and fishery workers | 6 (17) | 8 (23) | 10 (29) | |
| Service and sales worker | 4 (11) | 1 (3) | 2 (6) | |
| Elementary occupation | 2 (6) | 0 (0) | 3 (9) | |
| Craft related trades workers | 2 (6) | 0 (0) | 2 (6) | |
| Professional | 1 (3) | 2 (6) | 0 (0) | |
| Technicians and associate professionals | 0 (0) | 1 (3) | 0 (0) | |
| Clerical support work | 0 (0) | 1 (3) | 0 (0) | |
| Other | 1 (3) | 2 (6) | 0 (0) | |
| Accompanying person, n (%) | ||||
| Preoperative appointment | 30 (86) | 31 (89) | 26 (74) | 0.25 |
| Day 1 postoperative appointment | 33 (94) | 30 (86) | 27 (77) | 0.12 |
| Month 1 postoperative appointment | 16 (46) | 20 (57) | 24 (69) | 0.15 |
| Proportion of public transport against private, n (%) | ||||
| Preoperative appointment | 23 (66) | 27 (77) | 24 (69) | 0.64 |
| Day 1 postoperative appointment | 23 (66) | 17 (49) | 24 (69) | 0.08 |
| Month 1 postoperative appointment | 25 (71) | 28 (80) | 30 (86) | 0.38 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; significant P values are in bold; aP values calculated using the Kruskal–Wallis test for continuous variables and Fisher’s exact test for categorical variables; bUnits in kilometers (km); cUnits in Indian Rupees (INR); dEarnings exclusively from active earners (excluding those with no occupation, retired, or on pension). Units in Rupees (INR)
Travel burden
Table 2 shows the reported travel burden. For preoperative appointments, VC-only had the shortest transport time (66 ± 41 min; UBH-only: 198 ± 140 min; UBH/VC: 258 ± 106 min; P < 0.001). For the POD1 appointment, UBH/VC had the shortest travel time (54 ± 41 min; UBH-only: 195 ± 143 min; VC-only: 66 ± 41 min; P < 0.001). For the POM1 appointments, UBH-only had the shortest travel time (195 ± 137 min; UBH/VC: 256 ± 100 min; VC-only: 221 ± 87 min; P = 0.02). When examining the entire appointment period (i.e. preoperative, POD1, and POM1 visits), transport time was shortest in VC-only (353 ± 118 min; UBH-only: 589 ± 418 min; UBH/VC: 568 ± 230 min; P < 0.001).
Table 2.
Travel burden
| UBH-only (n=35) | UBH/VC (n=35) | VC-only (n=35) | P a | |
|---|---|---|---|---|
| Transport time, mean (SD)b | ||||
| Preoperative appointment | 198 (140) | 258 (106) | 66 (41) | <0.001 |
| Day 1 postoperative appointment | 195 (143) | 54 (41) | 66 (41) | <0.001 |
| Month 1 postoperative appointment | 195 (137) | 256 (100) | 221 (87) | 0.02 |
| Entire appointment period | 589 (418) | 568 (230) | 353 (118) | <0.001 |
| Total appointment time, mean (SD)c | ||||
| Preoperative appointment | 497 (171) | 564 (235) | 131 (63) | <0.001 |
| Day 1 postoperative appointment | 294 (179) | 110 (63) | 127 (61) | <0.001 |
| Month 1 postoperative appointment | 278 (140) | 346 (144) | 304 (141) | <0.05 |
| Entire appointment period | 1069 (439) | 1021 (383) | 562 (177) | <0.001 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; aP value calculated using the Kruskal–Wallis test for continuous variables and Fisher’s exact test; Significant P values are in bold; bTransportation time for round trip to hospital (units in minutes); cTotal appointment time for home–hospital–home trip (units in minutes)
The total appointment time analysis showed that VC-only had the shortest preoperative appointment (131 ± 63 min; UBH-only: 497 ± 171 min; UBH/VC: 564 ± 235 min; P < 0.0011). For the POD1 appointment, UBH/VC had the shortest total appointment time (110 ± 63 min; UBH-only: 294 ± 179 min; VC-only: 127 ± 61 min; P < 0.001). For the POM1 appointment, UBH-only had the shortest total appointment time (278 ± 140 min; VC-only: 304 ± 141 min; UBH/VC: 346 ± 144 min; P < 0.05). Over the entire appointment period, the total appointment time was lowest in VC-only (562 ± 177 min; UBH-only: 1069 ± 439 min; UBH/VC: 1021 ± 383 min; P < 0.001).
VC-only consistently demonstrated significant reductions in transport and total appointment times compared to UBH-only. For the preoperative appointment, VC-only reduced transport time by 132 min (95% CI: −182 to −83, P < 0.001) and total appointment time by 367 min (95% CI: −448 to −285, P < 0.001). On POD1, transport time decreased by 129 min (95% CI: −171 to −87, P < 0.001) and total appointment time reduced by 166 min (95% CI: −221 to −112, P < 0.001). Although no significant reduction was observed for POM1 transport or total appointment time (P = 0.33 and P = 0.78, respectively), the cumulative effect over the entire appointment period was substantial. VC-only achieved a total reduction of 236 min in transport time (95% CI: −371 to −102, P = 0.001) and 507 min in total appointment time (95% CI: −673 to −340, P < 0.001) compared to UBH-only. In contrast, UBH/VC showed mixed results, with significant reductions in transport and total appointment times on POD1 (141 min and 183 min, respectively; P < 0.001) but increased travel burdens for POM1 appointments and no significant cumulative reductions over the entire period (P = 0.76 for transport time; P = 0.57 for total appointment time) [Table 3].
Table 3.
Linear regression model coefficients for travel burden
| ß (95% CI) |
P a | ß (95% CI) | P a | ||
|---|---|---|---|---|---|
| UBH-only (n=35) | UBH/VC (n=35) | VC-only (n=35) | |||
| Transport time, meanb | |||||
| Pre-operative appointment | Ref | 60 (11 to 109) | 0.02 | −132 (−182 to−83) | <0.001 |
| Day 1 post-operative appointment | Ref | −141 (−184 to−99) | <0.001 | −129 (−171 to−87) | <0.001 |
| Month 1 post-operative appointment | Ref | 61 (8 to 113) | 0.02 | 26 (−27 to 78) | 0.33 |
| Entire appointment period | Ref | −21 (−155 to 114) | 0.76 | −236 (−371 to−102) | 0.001 |
| Total appointment time, meanc | |||||
| Pre-operative appointment | Ref | 67 (−14 to 149) | 0.11 | −367 (−448 to−285) | <0.001 |
| Day 1 post-operative appointment | Ref | −183 (−238 to−129) | <0.001 | −166 (−221 to−112) | <0.001 |
| Month 1 post-operative appointment | Ref | 68 (1 to 135) | <0.05 | 26 (−41 to 93) | 0.44 |
| Entire appointment period | Ref | −48 (−215 to 118) | 0.57 | −507 (−673 to 340) | <0.001 |
CI=confidence interval, UBH=urban base hospital, VC=vision center; aP value calculated using a simple linear regression model; Significant P values are in bold; bTransportation time for round trip to hospital (units in minutes); cTotal appointment time for home–hospital–home trip (units in minutes)
Financial burden (missed wages)
Table 1 presents the occupational distribution. The proportion of occupation categories were similar across groups (P = 0.14). Among earning participants, the most common occupation was “Skilled farm and fishery workers” (UBH-only: 17%; UBH/VC: 23%; VC-only: 29%). Adjusted wages, accounting solely for earning participants, excluded patients in the occupation category of “Retired/Pensioner” and “No occupation,” and were significantly higher for UBH/VC (1115 ± 1167; UBH-only: 561 ± 388; VC-only: 294 ± 296 INR, P < 0.01). This indicates a notable disparity in wages across the groups, with UBH/VC demonstrating the highest earning potential.
Supplementary Table 1 displays missed work occurrences across the groups throughout the appointments. The mean missed wages were lower in VC-only for all visits, as no active earners reported missing work to attend their appointments at the VC. The impact of VCs was also evident for the POD1 appointment, during which the UBH/VC attended the VC, resulting in a reduction of 238 rupees in all missed wages and 60 rupees in adjusted missed wages compared to the missed wages of UBH-only [Table 4]. No significant differences were observed in missed wages among caretakers, possibly due to the limited sample size [Supplementary Table 2].
Supplementary Table 1.
Number of participants who missed work due to appointment
| UBH-only (n=35) | UBH/VC (n=35) | VC-only (n=35) | Total Participants (n=105) | P c | |
|---|---|---|---|---|---|
| Preoperative appointment, mean (SD) | |||||
| Participants from all occupations, n (%)a | 11 (31) | 13 (37) | 1 (3) | 25 (24) | 0.001 |
| Participants from adjusted occupations, n (%)b | 9 (44) | 12 (34) | 0 (0) | 21 (20) | <0.001 |
| Day 1 postoperative appointment, mean (SD) | |||||
| Participants from all occupations, n (%)a | 11 (31) | 10 (29) | 1 (3) | 22 (21) | 0.005 |
| Participants from adjusted occupations, n (%)b | 9 (44) | 8 (23) | 0 (0) | 17 (16) | 0.001 |
| Month 1 postoperative appointment, mean (SD) | |||||
| Participants from all occupations, n (%)a | 5 (14) | 8 (23) | 0 (0) | 13 (10) | 0.01 |
| Participants from adjusted occupations, n (%)b | 4 (25) | 6 (40) | 0 (0) | 10 (10) | 0.02 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; Significant P values are shown in bold; aFor “participants from all occupations,” each group (UBH-only, UBH/VC, VH-only) included 35 participants; bFor “participants from adjusted occupations,” the sample included 16 participants in UBH-only, 15 in UBH/VC, and 17 in VC-only; cP value calculated using a Chi-squared test
Table 4.
Financial burden (missed wage of the participant)a
| UBH-onlyb | UBH/VCb | VC-onlyb | P c | |
|---|---|---|---|---|
| Preoperative appointment, mean (SD) | ||||
| All missed wage | 374 (1340) | 463 (941) | 0 (0) | 0.03 |
| Adjusted missed wage | 324 (351) | 1081 (1198) | 0 (0) | <0.001 |
| Day 1 postoperative appointment, mean (SD) | ||||
| All missed wage | 394 (1343) | 156 (421) | 0 (0) | 0.06 |
| Adjusted missed wage | 305 (349) | 365 (590) | 0 (0) | 0.01 |
| Month 1 postoperative appointment, mean (SD) | ||||
| All missed wage | 106 (292) | 145 (422) | 0 (0) | 0.42 |
| Adjusted missed wage | 169 (350) | 338 (602) | 0 (0) | 0.14 |
| Entire appointment period, mean (SD) | ||||
| All missed wage | 873 (2721) | 765 (1452) | 0 (0) | 0.01 |
| Adjusted missed wage | 798 (988) | 1784 (1781) | 0 (0) | <0.001 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; aunits in Indian Rupees (INR); bFor “all missed wage,” UBH-only, UBH/VC, and VC-only include missed wage data from 35 participants each. For “adjusted missed wage,” which excludes missed wages from individuals in the “Retired/On pension” or “No occupation” categories, complete data were available for 16 UBH-only participants, 15 UBH/VC participants, and 17 VC-only participants; cP value calculated using the Kruskal–Wallis test; Significant P values are in bold
Supplementary Table 2.
Financial burden (missed wage of the caretaker)a
| UBH-only (n=17) | UBH/VC (n=10) | VC-only (n=7) | P b | |
|---|---|---|---|---|
| Preoperative appointment, mean (SD) | 1541 (2448) | 937 (1008) | 679 (511) | 0.58 |
|
| ||||
| UBH-only (n=18) | UBH/VC (n=10) | VC-only (n=8) | P | |
|
| ||||
| Day 1 postoperative appointment, mean (SD) | 1261 (2476) | 2147 (3370) | 694 (475) | 0.62 |
|
| ||||
| UBH-only (n=6) | UBH/VC (n=7) | VC-only (n=8) | P | |
|
| ||||
| Month 1 postoperative appointment, mean (SD) | 1967 (3952) | 1080 (1229) | 1844 (3329) | 0.59 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; aUnits in Rupees (INR); bP value calculated using the Kruskal–Wallis test
Financial burden (Transportation cost)
VC-only consistently incurred the lowest transportation costs for all appointments. These differences were statistically significant for the preoperative (44 ± 26 INR; P < 0.001) and POD1 appointments (44 ± 26 INR; P < 0.001), but not for the POM1 appointment (163 ± 38 INR; P = 0.16). UBH-only had the highest transportation costs across all appointments, followed by UBH/VC, which showed intermediate costs. The mean transportation costs for UBH/VC and VC-only at POM1 were similar (195 vs. 163 INR; P = 0.16), but UBH/VC had a larger SD (309 vs. 38 INR) [Table 5].
Table 5.
Financial burden (transportation cost)a
| UBH-only (n=35) | UBH/VC (n=35) | VC-only (P=35) | P b | |
|---|---|---|---|---|
| Preoperative appointment, mean (SD) | 532 (1058) | 307 (545) | 44 (26) | <0.001 |
|
| ||||
| UBH-only (n=35) | UBH/VC (n=30) | VC-only (n=35) | P | |
|
| ||||
| Day 1 postoperative appointment, mean (SD) | 544 (1055) | 103 (188) | 44 (26) | <0.001 |
|
| ||||
| UBH-only (n=34) | UBH/VC (n=32) | VC-only (n=34) | P | |
|
| ||||
| Month 1 postoperative appointment, mean (SD) | 499 (1068) | 195 (309) | 163 (38) | 0.16 |
SD=standard deviation, UBH=urban base hospital, VC=vision center; Significant P values are shown in bold; aUnits in Rupees (INR); bP value calculated using the Kruskal–Wallis test
Participants’ perceptions of decentralized follow-up
63% of UBH/VC participants (n = 22) preferred attending postoperative appointments at VCs, while 37% (n = 13) preferred the UBH. Participants who preferred VCs cited its proximity to their homes, while those who preferred the UBH valued the ability to see a doctor in person.
Discussion
Our survey reveals significant travel and financial burdens for cataract patients and caretakers in rural, low-resource settings, and suggests that decentralized cataract care may provide a patient-centered solution to key barriers in LMICs.
Reduced transportation and total appointment time
Patients undergoing all follow-up at VCs reported >3 h reduction in transportation time compared to those at UBH (353 ± 118 vs. 589 ± 418 min) and UBH/VC (568 ± 230 min; P < 0.001) for the entire appointment period. Similarly, total appointment time for VC-only (562 ± 177 min) was >7 h shorter than for UBH-only (1069 ± 439 min) and UBH/VC (1021 ± 383 min; P < 0.001), indicating a severe travel burden associated with travel to UBH [Table 2]. Linear regression confirmed these reductions, with VC-only showing the largest decreases in transport time (−236 min, 95% CI: −371 to −102, P = 0.001) and total appointment time (−507 min, 95% CI: −673 to −340, P < 0.001) compared to UBH-only. Notably, UBH/VC also experienced a significant reduction in POD1 transport time (−141 min, 95% CI: −184 to −99, P < 0.001), underscoring the impact of VCs even for partial follow-ups [Table 3]. These findings suggest that integrating VCs into care models offers meaningful reductions in travel burden and improves accessibility in rural and low-resource settings.
Decentralized care was positively perceived by patients, with 63% of UBH/VC participants preferring VCs for their appointment location, citing proximity to homes as the main factor. Reducing travel time via VCs may be an effective strategy for addressing the healthcare access inequities faced by people living in LMICs, where nearly half of the world’s population lacks access to essential health services due to geographic barriers.[13,14]
Reduced transportation costs from local follow-up
Local VC follow-up was associated with marked reductions in patients’ transportation costs, suggesting that decentralized follow-up could reduce financial barriers faced by low-resource populations when accessing cataract surgical care. While there was no significant difference in the use of public versus private transport across the entire cohort [Table 1], the mean transportation cost was consistently lowest in VC-only, with significant differences observed compared to UBH-only and UBH/VC for preoperative and POD1 appointments (P < 0.001). At POM1, the mean transportation costs were similar between UBH/VC and VC-only (195 vs. 163 INR). However, UBH/VC had a significantly larger SD compared to VC-only (309 vs. 38 INR), indicating the variability likely contributed to the nonsignificant P value (0.16). The largest reduction was seen in POD1 appointment costs, where VC-only spent 500 rupees less on transport (44 ± 26 INR) compared to UBH-only (544 ± 1055 INR; P < 0.001) [Table 5]. For perspective, the average daily wage for a working-age rural farmer and primary breadwinner in Tamil Nadu was 360 INR,[15] meaning a decentralized appointment at a local VC would save a rural family one day’s wages from transportation cost savings alone.
Missed wages from appointments
The survey evaluated missed wages as a measure of the financial burden associated with attending appointments. Significant differences were observed in adjusted missed wages across the three groups, with no participants in the VC-only group missing wages to attend their appointment. Over the entire appointment period, adjusted missed wages were highest in UBH/VC (1784 ± 1781 INR) compared to UBH-only (798 ± 988 INR, P < 0.001), while no missed wages were reported in VC-only [Table 4]. These findings offer two insights. First, the mixed-appointment location model may not provide the same level of wage savings as exclusively attending a VC. Second, participants in UBH/VC may have experienced a greater economic impact from missed work, as evidenced by their higher daily wages [Table 1], which likely motivated their decision to opt for decentralized follow-ups closer to home to minimize work absenteeism.
Decentralized follow-up: The ideal versus the real
Ophthalmologists at urban centers may prefer rural patients to follow up in person. However, the global shortage of ophthalmologists, their geographic clustering in urban centers, and the travel and financial burdens faced by low-resource populations seeking surgical care often result in poor follow-up and worse outcomes, emphasizing the need for patient-centered care models.[16,17,18]
Financial and travel barriers negatively impact postoperative outcomes, particularly in LMICs, where nearly 5 billion people face limited access to surgery.[19] High transport costs and poor infrastructure lead to missed or delayed appointments, increasing the risk of poor postoperative outcomes.[14,20] Olson et al.[21] reported that travel burdens influenced hospital costs, length of stay, and recovery outcomes for urologic procedures. O’Connor et al.[22] linked longer travel distances for biliary cancer surgeries to lower household incomes and reduced survival rates. Future research should investigate how travel and financial burdens contribute to loss to follow-up among low-resource populations undergoing ophthalmic surgery.
A key question for decentralized cataract care models is whether they provide the same quality of care as urban-based follow-ups. Aravind employs a highly skilled workforce of para-professional staff proficient in anterior segment slit lamp examination, supported by live telecommunication with the ophthalmologists at UBH to clarify examination findings and rule out postoperative complications. To maintain safety, the threshold for referring rural patients with suspected postoperative complications to the urban tertiary facility should remain low. Further research should compare the safety profile of patients following up at VCs or UBH, either via adequately powered retrospective outcome analyses or, more ideally, through a prospective randomized trial. Such analyses would provide higher-quality information regarding the safety profile of decentralized surgical care before scaling it to other healthcare contexts.
The World Health Organization emphasizes seven pillars of quality of care, including effectiveness, safety, people-centeredness, timeliness, equity, integration, and efficiency in health services.[23,24] A decentralized care model may enhance effectiveness, people-centeredness, timeliness, and efficiency in healthcare delivery. While decentralized surgical follow-up could offer a more equitable approach to follow-up care, this model requires further validation. Research is needed to rigorously compare the safety profiles, clinical outcomes, cost-effectiveness, and environmental impact of traditional versus decentralized cataract surgical care across different geographic, epidemiologic, and economic contexts. Future research should also evaluate the cost-effectiveness of decentralized cataract care from patient and caretaker perspectives and assess whether reducing financial barriers improves follow-up and prevents complications.
Limitations
This study has limitations. Its nonrandomized, descriptive design limits causal inferences between decentralized care and reductions in travel time or costs. Nonetheless, the nonrandomized approach aimed to reflect real-world clinical conditions by recruiting patients based on actual attendance, ensuring representation from both decentralized and centralized care models while addressing logistical constraints. Participants choosing VCs had longer initial travel times to the UBH, likely biasing results away from the null. The use of means and SD was intended to capture the full range of responses—including outliers—to highlight public health relevance, particularly for individuals facing disproportionate burdens. However, this approach, combined with the limited sample size and single-center South Indian setting, may reduce statistical power and generalizability. However, our findings may have global relevance, as the VC care model is increasingly employed in other LMICs.[11] Larger prospective randomized studies of decentralized follow-up care models across diverse contexts could address these limitations. This study did not capture all burdens of cataract care, excluding costs such as accommodation, food, and drivers’ expenses. Additionally, in cases with multiple caretakers or accompanying persons, financial impacts may have been underestimated. Psychological (e.g., travel fatigue), social (e.g., missed education), and environmental (e.g., carbon footprint) factors warrant further investigation.
Conclusions
Cataract surgery patients undergoing follow-up at nearby technician-run VCs reported significantly reduced travel time and cost compared to patients following up at an urban eye hospital. A complete set of perioperative appointments conducted at VCs was associated with a significantly lower overall travel and financial burden, including reduced transport times, travel costs, and missed wages. This decentralized care delivery model may be more patient-centered and could help overcome travel- and cost-associated barriers to accessing cataract surgical care among low-resource populations.
Ethical committee approval
This study was approved by the Institutional Review Boards (IRBs) of Johns Hopkins Bloomberg School of Public Health and Aravind Eye Care System.
Conflicts of interest:
There are no conflicts of interest.
Funding Statement
Nil.
References
- 1.Lou L, Ye X, Xu P, Wang J, Xu Y, Jin K, et al. Association of sex with the global burden of cataract. JAMA Ophthalmol. 2018;136:116–21. doi: 10.1001/jamaophthalmol.2017.5668. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Fang Z, Chen XY, Lou LX, Yao K. Socio-economic disparity in visual impairment from cataract. Int J Ophthalmol. 2021;14:1310–4. doi: 10.18240/ijo.2021.09.03. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wang W, Yan W, Fotis K, Prasad NM, Lansingh V, Taylor HR, et al. Cataract surgical rate and socioeconomics: A global study. Invest Ophthalmol Vis Sci. 2016;57:5872–81. doi: 10.1167/iovs.16-19894. [DOI] [PubMed] [Google Scholar]
- 4.The International Agency for the Prevention of Blindness (IAPB) United Nations General Assembly Resolution on Vision. Published 2021 Available from: https://www.iapb.org/advocate/eye-health-and-sdgs/united-nations-general-assembly-resolution-on-vision/ . [Last accessed on 2024 Jan 31]
- 5.The International Agency for the Prevention of Blindness (IAPB) Whoever, wherever, whenever: Improving access to quality eye care for forcibly displaced populations. Published 2021 Available from: https://www.iapb.org/blog/whoever-wherever-whenever-improving-access-to-quality-eye-care-for-forcibly-displaced-populations/ . [Last accessed on 2024 Jan 31]
- 6.Rao GN, Khanna R, Payal A. The global burden of cataract. Curr Opin Ophthalmol. 2011;22:4–9. doi: 10.1097/ICU.0b013e3283414fc8. [DOI] [PubMed] [Google Scholar]
- 7.Ali M, Awidi A, Wang J, Varadaraj V, Cai C, Srikumaran D, et al. The association of social determinants of health with cataracts in the U.S. Investig Ophthalmol Vis Sci. 2022;63:4243. [Google Scholar]
- 8.Vajpayee RB, Joshi S, Saxena R, Gupta SK. Epidemiology of cataract in India: Combating plans and strategies. Ophthalmic Res. 1999;31:86–92. doi: 10.1159/000055518. [DOI] [PubMed] [Google Scholar]
- 9.The Hindu Nearly 5 million people in India internally displaced due to climate change, disasters in 2021: UN. Available from: https://www.thehindu.com/news/national/nearly-5-million-people-in-india-internally-displaced-due-to-climate-change-disasters-in-2021-un/article65535820.ece . Published June 17, 2022. [Last accessed on 2024 Jan 31]
- 10.Aravind Eye Care System Aravind Eye Care System. Our Story. Published 2022 Available from: https://aravind.org/our-story/ . [Last accessed on 2024 Jan 31]
- 11.Muralikrishnan J, Christy J, Srinivasan K, Subburaman G, Shukla A, Venkatesh R, et al. Access to eye care during the COVID-19 pandemic, India. Bull World Health Organ. 2022;100:135–43. doi: 10.2471/BLT.21.286368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Aravind Eye Care system Aravind Eye Care System. Vision Centres. Published 2022 Available from: https://aravind.org/vision-centre/ . [Last accessed on 2024 Jan 31]
- 13.World Health Organization World Bank and WHO: Half the world lacks access to essential health services-100 million still pushed into extreme poverty because of health expenses. Published December 13, 2017 Available from: https://www.who.int/news/item/13-12-2017-world-bank-and-who-half-the-world-lacks-access-to-essential-health-services-100-million-still-pushed-into-extreme-poverty-because-of-health-expenses . [Last accessed on 2024 Jan 31]
- 14.Dawkins B, Renwick C, Ensor T, Shinkins B, Jayne D, Meads D. What factors affect patients’ ability to access healthcare? An overview of systematic reviews. Trop Med Int Health. 2021;26:1177–88. doi: 10.1111/tmi.13651. [DOI] [PubMed] [Google Scholar]
- 15.CEIC Data Average Daily Wage Rate: Rural: Agricultural: Packaging and Agriculture: Men: Tamil Nadu. Available from: https://www.ceicdata.com/en/india/average-daily-wage-rate-rural-agricultural-by-state-packaging-and-agriculture/average-daily-wage-rate-rural-agricultural-packaging-and-agriculture-men-tamil-nadu . Published 2019. [Last accessed on 2024 Jan 31]
- 16.Resnikoff S, Felch W, Gauthier TM, Spivey B. The number of ophthalmologists in practice and training worldwide: A growing gap despite more than 200 000 practitioners. Br J Ophthalmol. 2012;96:783–7. doi: 10.1136/bjophthalmol-2011-301378. [DOI] [PubMed] [Google Scholar]
- 17.Berkowitz ST, Finn AP, Parikh R, Kuriyan AE, Patel S. Ophthalmology workforce projections in the United States, 2020 to 2035. Ophthalmology. 2024;131:133–9. doi: 10.1016/j.ophtha.2023.09.018. [DOI] [PubMed] [Google Scholar]
- 18.Weinhold I, Gurtner S. Understanding shortages of sufficient health care in rural areas. Health Policy (New York) 2014;118:201–14. doi: 10.1016/j.healthpol.2014.07.018. [DOI] [PubMed] [Google Scholar]
- 19.Varela C, Young S, Mkandawire N, Groen RS, Banza L, Viste A. Transportation barriers to access health care for surgical conditions in Malawi a cross sectional nationwide household survey. BMC Public Health. 2019;19:264. doi: 10.1186/s12889-019-6577-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Syed ST, Gerber BS, Sharp LK. Traveling towards disease: Transportation barriers to health care access. J Community Health. 2013;38:976–93. doi: 10.1007/s10900-013-9681-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Olson DJ, Gore JL, Daratha KB, Roberts KP. Travel burden and the direct medical costs of urologic surgery. J Health Econ Outcomes Res. 2016;4:47–54. doi: 10.36469/9825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.O’Connor SC, Mogal H, Russell G, Ethun C, Fields RC, Jin L, et al. The effects of travel burden on outcomes after resection of extrahepatic biliary malignancies: Results from the US extrahepatic biliary consortium. J Gastrointest Surg. 2017;21:2016–24. doi: 10.1007/s11605-017-3537-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ting DSJ, Buchan JC. Equity, access, and carbon cost-effectiveness of bilateral cataract surgery. Lancet. 2024;403:353–4. doi: 10.1016/S0140-6736(23)01923-2. [DOI] [PubMed] [Google Scholar]
- 24.World Health Organization Quality of Care. Published 2024 Available from: https://www.who.int/health-topics/quality-of-care#tab=tab_1 . [Last accessed on 2024 Feb 06]