Figure. Topics discussed in focus groups of physicians, patients, and village health workers (VHWs) in 3 counties in rural Guangdong Province, China. DR indicates diabetic retinopathy.
Yan X, Liu T, Gruber L, He M, Congdon N. Attitudes of Physicians, Patients, and Village Health Workers Toward Glaucoma and Diabetic Retinopathy in Rural ChinaA Focus Group Study. Arch Ophthalmol. 2012;130(6):761-770. doi:10.1001/archophthalmol.2012.145
Author Affiliations: Zhongshan Ophthalmic Center, Division of Preventive Ophthalmology, and State Key Laboratory, Sun Yat Sen University, Guangzhou, China (Drs Yan, He, and Congdon and Mss Liu and Gruber); and ORBIS International, New York, New York (Dr Congdon).
Objective To understand the knowledge and attitudes of rural Chinese physicians, patients, and village health workers (VHWs) toward diabetic eye disease and glaucoma.
Methods Focus groups for each of the 3 stakeholders were conducted in 3 counties (9 groups). The focus groups were recorded, transcribed, and coded using specialized software. Responses to questions about barriers to compliance and interventions to remove these barriers were also ranked and scored.
Results Among 22 physicians, 23 patients, and 25 VHWs, knowledge about diabetic eye disease was generally good, but physicians and patients understood glaucoma only as an acutely symptomatic disease of relatively low prevalence. Physicians did not favor routine pupillary dilation to detect asymptomatic disease, expressing concerns about workflow and danger and inconvenience to patients. Providers believed that cost was the main barrier to patient compliance, whereas patients ranked poorly trained physicians as more important. All 3 stakeholder groups ranked financial interventions to improve compliance (eg, direct payment, lotteries, and contracts) low and preferred patient education and telephone contact by nurses. All the groups somewhat doubted the ability of VHWs to screen for eye disease accurately, but patients were generally willing to pay for VHW screening. The VHWs were uncertain about the value of eye care training but might accept it if accompanied by equipment. They did not rank payment for screening services as important.
Conclusions Misconceptions about glaucoma's asymptomatic nature and an unwillingness to routinely examine asymptomatic patients must be addressed in training programs. Home contact by nurses and patient education may be the most appropriate interventions to improve compliance.
Glaucoma and diabetic retinopathy (DR) are among the leading causes of blindness in the world1 and in China.2,3 Glaucoma was estimated to affect 60.5 million persons in 2010,4 and it seems that the prevalence of diabetes in the world's most populous countries is growing rapidly,5,6 likely an indication of a commensurate rise in associated eye disease. Glaucoma and DR have been included as targets for reduction in the burden of avoidable blindness by Global Vision 2020,7 and each has therapies that have been proved effective in large randomized controlled trials.8,9
Nevertheless, the treatment of glaucoma and DR presents significant challenges in areas of limited resources compared with conditions such as age-related cataract. Glaucoma and DR are frequently asymptomatic until substantial, often irreversible damage to vision has already occurred. Compared with cataract surgery, a 1-time procedure that yields vision improvement in approximately 90% of patients,10 treatments for glaucoma and DR may require long-term follow-up, do not commonly produce dramatic improvement in vision, and, in fact, may be associated with decreased central acuity,11 cataract,12 and compromise of peripheral and color vision.13 In part because of these complexities, access to care for these diseases remains poor: it has been estimated that only 10% of persons in India with glaucoma are diagnosed,14 and the Handan Eye Survey15 reports that only 10% of persons with diabetes in rural areas of China have received diabetic eye care.
Available evidence16- 18 suggests that use of eye care in diabetic persons may be poor in the developing world. Although evidence19,20 is sparse from developing regions, problems with adherence to glaucoma therapies are well documented in the developed world,21 as is less-than-optimal adherence to preferred practice patterns for glaucoma care among physicians.22
To provide effective care for glaucoma and DR in areas with limited resources, education of eye care providers and interventions to promote compliance among patients are needed. However, little published information exists regarding the knowledge and attitudes of physicians and patients toward glaucoma and DR in the developing world.17,18 As part of preparation for a program that will aim to improve the quality of existing care for glaucoma and DR in rural China, we convened focus groups (FGs) concerning these diseases among rural patients, physicians, and village health workers (VHWs).
Focus groups can be helpful for assessing barriers to desired behaviors that are not well measured by simpler survey questions and for understanding complex factors that underlie health disparities. They can be of particular benefit in the early stages of research as part of the process of hypothesis generation. Finally, as in the present case, FGs may be helpful in designing educational programs and interventions tailored to the knowledge and attitudes of particular groups. Focus groups are not designed to provide the large sample sizes that may be achieved using shorter instruments based on closed-ended questions17,18 and, in fact, are complementary to such quantitative research methods.
The aim of the present study was to address the following questions:
What is the understanding of caregivers and patients regarding symptoms and available treatments for glaucoma and DR?
What do patients and caregivers see as the principal barriers to compliance with care for glaucoma and DR?
What are the attitudes of patients and caregivers toward a variety of widely used interventions to enhance adherence of patients to recommended care and to promote best care practices among physicians?
What role can VHWs play in outreach efforts to enhance patient acceptance of eye care?
The Guangzhou Comprehensive Rural Eye Service Training (CREST) is a 5-year collaboration between ORBIS International and the Zhongshan Ophthalmic Center to provide ophthalmologists currently performing cataract surgery at rural county-level hospitals in Guangdong Province, southern China, with comprehensive ophthalmic training. A specific goal is for these physicians to be able to diagnose and treat glaucoma and DR. A key part of the project will also be outreach screening and initiatives to create patient demand for glaucoma and DR care. The present research was conducted to better characterize physician and patient knowledge and attitudes toward these diseases to design more effective physician training modules and interventions to enhance patient compliance. Study personnel obtained written informed consent from all the participants, the ethics committees at the Zhongshan Ophthamic Center fully approved the protocol, and the study followed the tenets of the Declaration of Helsinki throughout.
All the FGs were conducted at 3 rural, county-level hospitals participating in the CREST network or at village health clinics associated with 1 of the 3 hospitals. Boluo County People's Hospital (public) services a population of 850 000 with an average annual income of ¥5760 (US $890 ). The annual number of outpatient visits is 12 800; 520 cataract procedures are performed each year. Sihui County People's Hospital (public) has a catchment population of 400 000 persons, whose annual income is ¥6210 (US $955 ). Annual figures for outpatient visits and cataract operations are 18 000 and 330, respectively. Yangjiang Ophthalmic Hospital is a private facility servicing a population of approximately 2 370 000 persons with an annual mean income of ¥5564 (US $856 ). The facility services 12 400 outpatients annually and performs 2100 cataract operations each year.
Three key constituencies (physicians, patients, and VHWs) participated in separate FGs in each of the 3 counties (a total of 9 FGs). The intent was to include 7 to 9 participants in each group to maximize the potential for interaction while minimizing the possibility of some participants not having the opportunity to speak. The 3 participant groups consisted of 22 physicians, 23 patients, and 25 VHWs.
Study personnel selected ophthalmologists (n = 8, 36%); eye, ear, nose, and throat specialists (n = 9, 41%); and internists (n = 5, 23%) at random from personnel lists at each of the 3 hospitals. At least 1 physician with eye training and 1 internist was included in each group.
Interviewers randomly selected patients with a diagnosis of glaucoma (n = 3, 13%), diabetes (n = 13, 57%), or DR (n = 7, 30%) from registration lists at outpatient and inpatient ophthalmology; eye, ear, nose, and throat; and internal medicine clinics at the 3 hospitals. At least 1 patient with glaucoma and 1 with diabetes or DR participated in each group.
The VHWs were selected at random from villages (mean [SD] population, 2600 ; range, 800-5000) in the 3 counties serviced by the hospitals. All the participants in the FGs received a meal and mineral water and were reimbursed as required for travel up to ¥50 ($7). No other payment was given.
After 3 sessions of intensive training by an experienced FG researcher (N.C.), 3 native speakers of Cantonese dialect were chosen as facilitators. They prepared a brief general introduction to the FG process and chose 1 or 2 open-ended questions on each of the topics listed in the Figure. The 3 facilitators prepared separate draft scripts for the physician, patient, and VHW FGs based on the introduction and questions as noted previously herein. All the coauthors reviewed the drafts, and an experienced FG researcher (N.C.) subsequently revised them.
Nine FGs were conducted between October 25, 2010, and November 18, 2010, one each for physicians, patients, and VHWs associated with each of the 3 county-level hospitals. The physician, patient, and VHW FGs lasted an average of 80, 60, and 70 minutes, respectively. All the FGs were conducted in Cantonese or Mandarin, recorded digitally, and subsequently transcribed into written Chinese by a trained team of 3 transcriptionists. Translations into English in the present article were performed by an ophthalmologist (N.C.) fluent in English and Chinese.
After completion of the FGs, participants in the physician, patient, and VHW groups were each asked to rank a variety of potential responses to 2 questions: “What are the most important barriers to patients coming to this hospital for treatment and follow-up?” and “Which of these methods of improving patients' demand for and adherence with care at this hospital is likely to be most effective?” Participants in the VHW group were additionally asked to rank a variety of potential types of training or equipment that might be provided and goals for health care in the village in response to the following question: “Which of the following is most important to you?” In addition to preselected responses, additional options were added at the time of each FG to accurately reflect the content of discussions. The rankings were performed anonymously.
The 3 FG facilitators coded all interview transcripts independently using qualitative data analysis software (NVivo 8.0; QSR International Inc) after independent review of all 9 transcripts and agreement on a coding scheme. The final scheme was entered as “tree nodes,” which were identical for both coders. The interrater coding reliability (Cohen κ) was calculated using the “coding comparison” query provided by NVivo as percentage agreement of passages coded to the appropriate nodes. By convention, a κ value greater than 0.7 is considered acceptable interrater reliability.23
Participants' ranked responses to questions, such as “What are the most important barriers to patients going to this hospital for treatment and follow-up?” were analyzed using a commercially available software program (SPSS, version 17.0; SPSS Inc). Where a total of x responses were to be ranked, x points were assigned for each participant who ranked the response highest, x −1 for a second-place ranking, and so forth, with a lowest ranking receiving a score of 1. A total score for each response was calculated as the sum of points awarded across all the participants, which was then divided by the maximum possible score to facilitate comparison across groups of different sizes. Thus, a maximum score was 1.0 and a minimum score was 1/ x.
The κ value comparing nodal coding between the 2 coders was 0.75. The mean (SD) age of the 22 physicians was 33 (4.8) years (age range, 25-44 years), 17 (77%) were men, and the mean (SD) working experience was 10.0 (4.8) years (range, 1-20 years). The mean age of the 23 patients was 61.8 (12.1) years (age range, 27-76 years), 14 (61%) were men, 15 (65%) had been diagnosed within 3 months, 11 (48%) had only a junior high school or less education, and 20 (87%) lived with family members. The visual acuity of patients was not recorded. The 25 VHWs had a mean (SD) age of 47 (12) years, 22 (88%) were men, their mean (SD) work experience was 24.5 (12) years (range, 3-43 years), and 21 (84%) had only a high school or vocational high school education.
Physicians and patients speaking during the FGs showed a good understanding of diabetes, including its pathologic characteristics, diagnosis, treatment, potential severity, and effect on vision. Physicians stated that diabetes was “a relative or complete insufficiency of insulin secretion, combined with insulin resistance” and noted that “complications may affect the eyes and wound healing on the feet.” There was wide understanding of the use of fasting blood glucose level to diagnose the disease and glycated hemoglobin level to monitor it. All the physician FGs could name a variety of diabetic ocular complications, including retinal vascular changes, growth of new blood vessels, lens opacity, and secondary glaucoma. Suggested techniques to diagnose eye disease included retinal examination, fundus photographs, and even optical coherence tomography. Patients tended to define diabetes using the popular Chinese phrase “3 excesses and 1 insufficiency” (eg, polyphagia, polydypsia, polyuria, and cachexia), although some could specifically state the role of insulin in the disease. Patients had a good awareness of the impact of diabetes on the eye, and several also indicated foot problems and coronary artery disease as potential complications.
Physicians identified a variety of treatments for diabetes focused on reduction of blood glucose levels. These treatments included dietary modification, exercise, medication, laser therapy for early DR, and vitrectomy for more advanced eye disease. The use of traditional Chinese medicine to “open the blood vessels” was also advocated in early disease. Some physicians indicated that treatment of DR “had comparatively poor results.” Physicians identified the importance of “early examination and reduction of blood sugar” and “examining the fundus as soon as the blood sugar is discovered to be high.” Several patients understood that diabetes required “taking medicine your whole life” and “is impossible to completely cure.” Significantly, patients did identify physicians as an important source of information about diabetes (along with “relatives” and “the newspaper”). Presumably related to their good understanding of disease, physicians widely agreed that diabetes and DR were common problems: “20% of hypertensives will have concurrent diabetes” and “many diabetics have at least stage I or II retinal disease.”
In contrast to diabetes, there were important and basic misconceptions about glaucoma pathogenesis, symptoms, detection, and, as a consequence, prevalence by physicians and patients. Physicians stated that “glaucoma is divided into open and closed angle, and the open-angle form is rarely seen,” “glaucoma results from ‘overripeness' of the lens,” “the pressure rises . . . and the cornea becomes cloudy,” and “chronic glaucoma is very rarely seen,” and they indicated that symptoms, such as photophobia, nausea, vomiting, vision loss, and conjunctival hyperemia, were the norm. Similarly, patients identified glaucoma as a symptomatic disease: “the main sign is headache and loss of vision” and “classically, it's nausea and vomiting.” Sources of knowledge about the disease included friends, relatives, and the newspaper but not physicians.
Although physicians could identify several means of diagnosing glaucoma (gonioscopy, visual field testing, intraocular pressure, corneal edema, and poor vision), none mentioned examination of the optic nerve. Many physicians relied on acute symptoms and slitlamp examination of the anterior chamber. Medical, laser, and incisional surgical treatments were listed by various physicians, with specific medications and operations being identified. However, only symptomatic glaucoma was recognized as an important cause of blindness: “acute glaucoma presenting spontaneously to clinic is very common.” Some patients recognized that glaucoma could be treated but indicated that “it's probably difficult,” “it takes a very long time,” and “it's very difficult to treat when it's severe.” Others indicated, “I've heard glaucoma is untreatable, you go blind.”
Physicians were asked specifically about the use of routine examination of the fundus with pupillary dilation to detect asymptomatic disease. Most physicians indicated that this was necessary only in certain patients: “you can first check the blood sugar, and if it's abnormal, then dilate the pupil,” “you can check the intraocular pressure first, then dilate the pupil.” Several others stressed that “if a glaucoma patient has high pressure, you can't dilate the pupil.” Specific barriers to routine dilation of the pupil included the following: “it will increase the doctors' workload,” “patients with common conditions like conjunctivitis won't accept dilation,” “rural patients arriving by motorcycle won't be willing to be dilated,” and patient cost.
Physicians and VHWs ranked cost as the most significant barrier to patient compliance (physician score, 0.865; VHW score, 0.933), and they also assigned high rankings to other financial barriers, such as lack of transportation (physician rank: No. 3 [score, 0.696]; VHW rank: No. 2 [score, 0.725]) and an accompanying family member (VHW rank: No. 3 [score, 0.65]) (Table 1 and Table 2). Physicians and VHWs assigned low ranks to poor training and poor service attitude among providers. Patients, on the other hand, ranked poor training among physicians as their second-most significant barrier (score, 0.724; the highest-ranked was “forgot about the appointment”), and they did not rank “cost” among their top 3 barriers (Table 3).
Physicians, patients, and VHWs were unanimous in assigning low ranks to financially based interventions: direct payments to physicians or patients, lotteries with participation limited to compliant patients, and contracts with rewards and fines based on adherence to follow-up schedules did not rank among the preferred 3 interventions for any of the stakeholder groups (Tables 4, 5, and 6). Physicians tended to object that such approaches “will strike people as commercial methods,” would be subject to abuse (“some people will come to the hospital for checkups to get money, even though they're not sick”), and would be effective in motivating only poor patients. Patients were skeptical about actually receiving payments or winning lotteries and felt that such methods were “too commercialized.”
Patient education (physician rank: No. 1 [score, 0.854]; VHW rank: No. 2 [score, 0.731]) and telephone calls by nurses to remind patients of appointments (physician rank: No. 2 [score, 0.775]; VHW rank: No. 1 [score, 0.783]; and patient rank: No. 1 [score, 0.805]) ranked as the preferred interventions to improve patient compliance among all 3 stakeholder groups (Tables 4, 5, and 6). Physicians and patients pointed out that an additional benefit of reminder telephone calls from nurses was that it “reflected concern” on the part of caregivers.
The VHWs were generally in agreement that their current level of training in eye care and diabetes is low: “we have no equipment” and “there are no screenings (for those conditions).” Except for infectious conditions, such as “red eye,” they generally believed that eye diseases, and DR in particular, were not common problems, but most agreed that these conditions were important because of their potential effect on vision.
The VHWs and physicians recognized that VHWs' location in the village gives them excellent local access to patients. However, both groups were concerned that patients might doubt the ability of VHWs to provide care for conditions, such as eye disease and diabetes, and might not readily pay for such care: “if they're being screened, they may not be willing to pay, as they don't feel they've had a real checkup.” Patients shared concerns about the competence of VHWs to conduct eye examinations and diabetic screening but were, in fact, generally willing to pay for their services: “wherever you go for an exam, you have to pay” and “I've always paid whatever they charged.” The VHWs ranked “being able to increase my earnings” last (rank No. 7; score, 0.344) when asked, “Which of the following is most important to you?” (Table 7).
The top 3 responses as ranked by VHWs in order of importance to them were “learning how to perform vision screenings” (score, 0.911), “learning how to screen for diabetes” (score, 0.766), and “receiving equipment for vision and diabetes screening” (score, 0.698) (Table 7). Many VHWs expressed concerns, however, including “limited time (for screenings),” “there will be little appeal (of these services) to the masses,” “it will take too much time to learn,” and “there are too few patients.” The need was stressed for training to be provided along with equipment, such as devices to measure blood glucose level and visual acuity.
A strong tendency was noted among participants in the present study to think of glaucoma solely as a dramatically symptomatic disease. This was compounded by physician unwillingness to perform routine dilated examinations on patients to detect nonsymptomatic conditions owing to a presumption of patient unwillingness, lack of time, unwarranted fears of harming patients with known glaucoma, and the view that focused examination of “at-risk” patients only would be sufficient to detect disease. In fact, population-based studies of glaucoma in China have indicated that open-angle glaucoma is more prevalent than closed-angle disease24,25 and that patients with intraocular pressure in the reference range (eg, ≤21 mm Hg) compose 83% to 85%24,26 of all glaucoma cases.
This combination of lack of knowledge and an unwillingness to look for nonsymptomatic disease presumably contributes to the perception among physicians that open-angle and chronic glaucoma are rare conditions. This is consistent with findings from elsewhere in the developing world that only 10% of prevalent glaucoma is diagnosed.14 A principal focus of CREST is to improve the diagnosis and treatment of glaucoma and DR in patients already presenting to the clinic not only through training but also by creating a standard of providing dilated examinations for all new patients. Results of the present study suggest that significant educational efforts directed toward physicians and patients may be required, as may initial subsidization of fees for such examinations.
The finding that knowledge of diabetes and diabetic eye disease was generally good among rural Chinese physicians and patients who spoke during the FGs may seem somewhat at odds with previous studies in China showing, for example, that only 41.6% to 55.3% of diagnosed diabetic patients in rural and urban clinics knew that diabetic persons required regular eye examinations.17 This points out the need for quantitative and qualitative research methods to obtain a more nuanced understanding of patient knowledge and attitudes about disease. Qualitative methods, such as the present study, may yield more in-depth knowledge but, owing to the format, where persons more knowledgeable on a particular topic may choose to speak, may not reveal areas of uncertainty as clearly as larger quantitative studies.
There has been much recent interest in various techniques to improve patient compliance with physician-recommended therapies27,28 and physician adherence to preferred practice patterns29 using financial interventions, such as direct payments, lotteries, and contracts mandating monetary rewards and penalties. Payments for performance for patients have been shown to be highly effective in promoting desired behaviors, such as weight loss30 and reduction in tobacco use,31 and may also be helpful in the control of chronic conditions, such as hypertension.32 In view of reports of poor adherence to recommended standards of care for diabetes16- 18 and poor compliance with recommended therapy for glaucoma19 in the developing world, effective interventions to improve compliance would be of substantial potential benefit. The results of the present study, however, suggest that physicians and patients in rural China may find such payment culturally unacceptable. Although similar strong reservations about these methods have been identified in studies of patient attitudes in the United States,33 the present results suggest that other methods, particularly patient education and telephone call reminders from nurses, may be better received in rural China. Such reminders and other strategies of “intensive case management” have also proved effective in improving compliance and outcomes in a variety of chronic diseases,34,35 although few such studies have been performed in the developing world. The fact that patients were supportive of such interventions in the present study is an encouraging indicator that intensive case management approaches might be effective in this setting. More work is needed to determine whether nurses are equally supportive of these approaches and whether they are practical in this context.
The fact that patients did not rank cost among their most significant barriers to receiving care is consistent with much,17,36,37 but not all,38 previous work on willingness to pay for eye care in rural China. This is a complex topic, affected by recent changes in the health care system, such as the introduction of a national health insurance program for rural dwellers, the New Cooperative Medical System. This system covers up to 60% to 70% of the cost of inpatient treatments (which includes most eye procedures in China) at an annual premium of only several US dollars per year. More than 95% of eligible persons participate in the system in rural China. The results of the present study add to a growing body of knowledge supporting the idea that cost recovery can be an important part of sustainability planning for eye care in rural China, critically important if access to care for complicated diseases, such as glaucoma and DR, is to be extended. The finding that patients ranked concerns over quality of physician care as an important barrier is consistent with previous quantitative studies in rural China and further underscores the need for high-quality medical training in new programs aimed at glaucoma and DR.
A key part of the proposed CREST program, and of efforts to increase demand for health care services in the developing world generally, is outreach screening. The VHWs are well positioned to play a role in such activities in rural China as they are located in each village throughout the country, and they are a regularly used source of basic health care. Results of the present study suggest that all the key stakeholder groups, VHWs included, have some reservations about their ability to perform screening for diabetes and eye disease. Significant training would likely be needed; the VHWs participating in the present FGs indicated a willingness to receive such training, but this was predicated on their also being provided with basic equipment to support the screening process. They are concerned about there being sufficient patients in need of ocular and diabetic care to justify their investment of time in receiving the training. It may be hoped that training of VHWs can be synergetic with educational interventions aimed at physicians and patients to ensure that sufficient numbers of patients are diagnosed and induced to seek care sustainably to keep VHWs active in a strengthened and improved eye care network. They may play a role in vision and blood glucose screening and in postoperative care after laser and incisional surgery for glaucoma and DR. The issue of payment for the services of VHWs will have to be resolved, although the present results suggest that this may not be the most critical issue for VHWs themselves and that patients may, in fact, be willing to make such payments if services are perceived to be of value.
The results and implications of the present study must be viewed in the context of its limitations. Focus group research is designed to be used in small groups of patients and to be complementary to quantitative methods by allowing complex issues to be explored in greater depth. As such, the number of individuals from each of the stakeholder groups was small. This is particularly true for patients with glaucoma, due in large part to the various barriers mentioned previously herein to the correct diagnosis of chronic glaucoma. Due to the small number of included patients and the fact that only 3 counties in 1 region were represented, these results may be applied to broader areas with caution only.
A strength of the present study is involvement of 3 key stakeholder groups in an area in which a large intervention is planned, making the results highly relevant. Also, data were collected and analyzed according to a predetermined protocol that has been used by one of us (N.C.) in previous studies in the area.38,39 Finally, the present study provides among the first data from the region on physician and patient knowledge and attitudes toward 2 diseases of increasing importance in rural Asia, together with the views of 3 critical groups on how best to improve adherence with long-term care for these conditions.
Correspondence: Nathan Congdon, MD, MPH, Zhongshan Ophthalmic Center, Division of Preventive Ophthalmology, 54 S Xianlie Rd, Yuexiu District, Guangzhou, People's Republic of China 510060 (firstname.lastname@example.org).
Submitted for Publication: August 12, 2011; final revision received November 8, 2011; accepted December 14, 2011.
Author Contributions: Dr Congdon had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Basic Research Fund of the State Key Laboratory, Zhongshan Ophthalmic Center.