Since the initiation and rapid scale-up of China's National Free Antiretroviral Treatment Program in 2002, more than 80,000 patients received antiretroviral therapy (ART) by December 2009, leading to a significant reduction of mortality among its participants.1 However, the benefit of ART cannot be fully realized without consistent adherence tightly linked to regular follow-up attendance.2 Missed clinic appointments are common even in resource-rich settings and are associated with delayed ART initiation, emergence of antiretroviral (ARV) resistance, virologic failure, and increased mortality.2–6
Previous studies have documented factors including financial, geographic, educational, and mental health that may act as barriers to HIV-infected patients entering and adhering to HIV care.7,8 It has also been reported that follow-up and treatment outcome is worse among those infected with HIV via injection drug use (IDU) or sexual transmission compared with other routes.9,10 Brennan et al11 observed that close to 30% of patients in Johannesburg, South Africa, missed at least 1 appointment in the first 6 months after ART initiation, and this was associated with a significantly higher risk of death. Poor medical adherence has been described among HIV patients in other developing countries as well,12,13 but has not been previously evaluated in China.
In this analysis, we describe the prevalence and pattern of visits missed within the first 6 months among patients initiating ART in China, analyze the impact of missed early visits on mortality, and determine factors associated with appointment nonadherence and determine other predictors of long-term survival for patients on ART with established outpatient HIV care. We hypothesize that missing early visits within the first 6 months after initiating treatment predicts adverse clinical outcomes and could serve as an early warning indicator to trigger additional outreach effort.
This study is an open retrospective cohort review of medical records collected through China's National Free Antiretroviral Treatment Program and its database. China's National Free Antiretroviral Treatment Program has been described in detail elsewhere.14,15 In brief, patients were eligible to receive highly active antiretroviral therapy (HAART) if they had CD4 count <200 cells per cubic millimeter (expanded to <350 cells/mm3 in 2008), total lymphocyte count <1.2 × 109 cells/L, or World Health Organization stage 3 or 4 disease.15 All patients initiating HAART were entered into and followed through the national database. Follow-up visits were scheduled for each patient after ART initiation at 2 weeks, 1 month, 2 months, 3 months, 6 months, and then every 3 months thereafter. Case report forms were completed by local health care workers at each visit and forwarded to the China Centers for Disease Control and Prevention through DataFax (Clinical DataFax Systems, Hamilton, Ontario, Canada).
All patients starting ART and registered into the database from June 2002 to June 2009 were screened. Patients older than 18, who were ARV naive at enrollment and who survived and accessed services for at least 7.5 months after enrollment were included. We excluded patients with missing age information, uncertain ART initiation or termination date, whose ART initiation date fell outside of the predefined time period, who were ART experienced or treated through another organization at enrollment, or whose ART regimen was not recorded, changed within the initial 6 months (because this restarted the follow-up visit schedule) or consisted of mono-, dual-, or zidovudine-stavudine (d4T) combinations (Fig. 1).
For each patient, window periods were created around the expected follow-up visits at 2 weeks, 1 month, 2 months, 3 months, and 6 months after treatment initiation. The window period around each visit was defined as the midpoint from the last visit to the midpoint until the next visit. An appointment was missed if no visit occurred within the window period. For instance, the window period for the 1-month visit begins at day 21 (midpoint between 2 weeks and 1 month) and ends at day 45 (midpoint between 1 month and 2 months). An additional 1.5 months were added as an acceptable delay for the 6-month visit, giving a total interval of 7.5 months after HAART initiation. Early missed visits were defined as nonattendance of any of the 5 routine appointments during the first 6 months of treatment.
Covariates in the analyses were abstracted from the treatment database, including sociodemographic factors (age, gender, marital status), HIV transmission route (sexual, IDU, plasma donation), baseline symptoms of opportunistic infections (OIs), hemoglobin, liver enzyme and CD4 cell count, and ART regimen. Given the difficulty in definitively diagnosing OIs in rural settings, signs and symptoms collected at baseline checklist screening and physical exams served as surrogates and were grouped into 5 categories including constitutional (fever, lymphadenopathy), pulmonary (cough, dyspnea, chest pain, night sweats), gastrointestinal (nausea, vomiting, diarrhea), skin or mucosal (rash, thrush, oral hairy leukoplakia), and neurological (headache or visual changes).15 Date of death was ascertained based on patients' hospital record, report of patients' relatives, or by national residence registration record if death occurred outside hospitals.
Descriptive statistics were performed for baseline characteristics of our study population. Cox proportional hazard models were used to evaluate factors associated with mortality, with patients stratified by the number of missed visits as follows: 0, 1–2, and 3 or more. Patients who were still active or lost to follow-up as of June 30, 2009, were censored. Kaplan–Meier plots were used to evaluate the survival of patients stratified by missed visit status. Logistic regression was used to evaluate factors associated with appointment nonadherence. Univariate factors with P < 0.1 and additional factors predetermined to be clinically meaningful based on expert opinion or the literature were included into full multivariate regression model. Data were analyzed using SAS version 9.1 (SAS Institute, Cary NC). All hypothesis testing was 2-sided, with α level of 0.05. This study was approved by the institutional review board of the Chinese Center for Disease Control and Prevention.
From 2002 to June 30, 2009, 68,364 patients initiated ART, of which 27,504 (40%) met the inclusion criteria (Fig. 1). Of note, the majority of patients (37,040) were excluded due to either ART initiation less than 6 months from the end of study period (18,766) or missing early follow-up data (18,274). Information on early follow-up was lost for many patients due to a system conversion in June 2006, which involved all patients enrolled in one province (Henan province) before this time. Sociodemographic and baseline characteristics of included patients are summarized in Table 1.
Including all enrolled patients, mean age (±SD) at ART initiation was 39 (±10) years and 16,821 (61%) were male. The most common risk factor for HIV transmission was plasma donation or blood transfusion (46%), followed by sexual transmission (33%) and IDU (22%). Median baseline CD4 (IQR) was 126 (44–205) cells per cubic millimeter at ART initiation. The median length of time on ART was 18.9 (interquartile range: 11.9–30.4) months. The most common first-line ARVs consisted of lamivudine (3TC), d4T, and nevirapine, used by 65%, 86%, and 82% of included patients, respectively.
Among the included patients, 37% missed at least one of the initial 5 appointments scheduled during the first 6 months, with 21%, 8%, 4%, 2%, and 2% missing 1, 2, 3, 4, and 5 visits, respectively. Appointment nonadherence rates for the first 5 scheduled visits were 14%, 14%, 14%, 12%, and 9.8%.
Factors Associated With Mortality
One thousand two hundred eighty (5%) patients died during the study period, with mortality significantly higher in patients with early missed visits (3.1 deaths per 100 person-years) compared with those who attended all scheduled visits (2.1 deaths per 100 person-years, P < 0.01). Figure 2 shows the step-wise effect of early missed appointments on mortality, with a hazard ratio: of 1.3 [95% confidence interval (CI) 1.1 to 1.5] for missing 1–2 visits and 1.7 (CI: 1.4 to 2.2) for missing 3 or more visits compared with no missed visits. Further analysis shows that missing visits during the first 3 months of treatment was also significantly associated with an increased risk of death (hazard ratio: 1.3, 95% CI: 1.1 to 1.5).
In multivariate Cox regression model (Table 2), higher mortality was also associated with older age, male gender, being not married, nonsexual HIV transmission, baseline CD4 <200 cells per cubic millimeter, baseline hemoglobin <80 g/L, having 1 or more baseline symptom categories, and being on a didanosine (ddI)-containing (vs. 3TC) or zidovudine-containing (vs. d4T) first-line regimen. Age ≥60 [adjusted hazard ratio (AHR), 3.4; CI: 2.4 to 5.0, compared with 18–29 of age], transmission through IDU (AHR, 2.3; CI: 1.9 to 2.8, compared with sexual transmission), and having 4–5 OI symptom categories at baseline (AHR: 2.2; CI: 1.7 to 2.8, compared with no baseline symptom categories) were the strongest predictors of mortality.
Factors Associated With Early Missed Visits
In multivariate logistic regression analyses, missed visits during the first 6 months of treatment were significantly more common among women [adjusted odds ratio (AOR): 1.1, CI: 1.0 to 1.3], patients with age ≥60 (AOR: 1.2, CI: 1.1 to 1.3, compared with age 18–29), those infected through IDU (AOR: 1.2, CI 1.1 to 1.3), or plasma donation or blood transfusion (AOR 1.3, CI: 1.2 to 1.4) compared with sexual transmission, those with baseline ALT at or above 100 U/L (AOR: 1.2, CI: 1.0 to 1.4), those with 2 or more baseline symptom categories (AOR: 1.2, CI: 1.1 to 1.3 for 2–3 categories; AOR: 1.5, CI: 1.3 to 1.6 for 4–5 categories, compared to no baseline symptoms), and those who received ddI-containing regimen instead of 3TC-containing regimen (AOR: 1.3, CI: 1.1 to 1.4). Having a baseline CD4 count of less than 200 cells per cubic millimeter (AOR: 0.8, CI: 0.7 to 0.9 for CD4 of 50–199 cells/mm3; AOR: 0.70, CI 0.6-0.8 for CD4<50 cells/mm3, compared with CD4 ≥200 cells/mm3) was protective. Nevirapine-containing versus an efavirenz (EFV)-containing first-line therapy and ART initiation after 2004 were associated with more missed early visits in the univariate but not multivariate analysis (Table 3).
As the ART program in China continues to scale-up in response to the expanding HIV epidemic, the importance of quality of care is increasingly being recognized. Our study is the first to explore the effect of missed early follow-up visits on HIV treatment and care in China, and our findings contribute to a growing body of evidence that demonstrate the impact of appointment adherence on patient survival, information which is especially limited in developing nations.
In this study, missed visits during the first 6 months after starting ART occurred in a large proportion of patients and were independently associated with increased mortality, with a dose-effect relationship. Among patients initiating ART, missing 1–2 of 5 visits in the first 6 months increased the risk of death by 30%, whereas missing 3–4 visits increased the risk by 70% (Table 2). Missed visits within the first 3 months of ART were associated with a 30% increase in mortality. The magnitude of the impact of missed visits on patients' long-term survival is comparable with that of baseline CD4 counts (Table 2). Patients often experience more side effects during the beginning months of ART and are more likely to develop drug resistance if nonadherent, given higher viral loads during this period. Our findings highlight the need for early interventions to improve appointment adherence and therefore long-term clinical outcomes among patients on HAART.16 Ma et al17 recently reported that patients treated at facilities below county level hospitals in China were more likely to experience virologic failure. Besides a higher proportion of nonphysician ART providers in the rural areas, as discussed by the authors, we hypothesize that higher rates of appointment nonadherence in rural areas may also have contributed to the higher failure rate.
We speculate that 2 major mechanisms link missed visits to increased mortality. First, nonattended visits may reflect medication nonadherence, which is associated with treatment failure and drug resistance. This assumption is supported by a similar dose-response relationship between appointment adherence and virologic response in HIV-infected patients undergoing treatment in other studies.3,18 A recent study from Kenya and Zambia found that even being late to a visit for 8 or more days was associated with a significantly higher risk of virologic failure (OR = 2.0).19 Second, those who have difficulty keeping scheduled appointments may have more advanced diseases. Identifying missed visits will help healthcare workers recognize and intervene earlier for patients at higher risk for poor treatment outcomes.
The solution to minimizing missed visits lies in understanding of its causes. In this study, those with more symptoms at baseline or elevated ALT were at the higher risk of missing visits (Table 3), possibly because poor physical health was a major barrier to traveling. Although transportation to clinic visits is funded through China's National Free Antiretroviral Treatment Program, commuting may be difficult or impossible for the very ill. ART-associated adverse effects may also partly explain the missed visits. Patients on a ddI-based first-line regimen, compared with a 3TC-based, missed more visits, likely because ddI is associated with more side effects. Consistent with this finding, the association between ddI-based regimens and higher rate of virologic failure was recently reported in Chinese patients.17 Worsening symptoms and side effects may create additional distrust in the health care system, and as such may be interpreted as inadequate care. Health care workers should be trained to manage side effects proactively. As recommend by World Health Organization,20 the more toxic first-line drugs, such as ddI, have been phased out of the national guidelines and replaced by 3TC. Focused efforts, including treatment, education and outreach, must be directed at patients presenting with multiple or severe symptoms to enhance appointment adherence.
It is interesting to note that even though more symptomatic disease at baseline was associated with more missed visits, lower baseline CD4 was protective against missed visits, after controlling for other factors (Table 3). The reasons for this contradiction are not clear, which deserve further studies. The association between lower baseline CD4 counts and higher rates of treatment success based on immunologic criteria was observed in China's National Free Antiretroviral Cohort.15 Other studies have reported that lower CD4 counts correlate with better retention in care, as patients with higher CD4 counts tend to be asymptomatic and therefore may be less motivated to remain in care.21 Moreover, patients with lower CD4 may receive more attention and counseling from health providers, which may facilitate their adherence.
Our study showed that Chinese patients who were infected with HIV through IDU and plasma donation missed more early visits compared with those infected through sexual transmission. Substance abuse has been linked to delays in accessing HIV care, poor adherence to medications, and to missed appointments in several studies,22–24 although more recent data from developed countries suggest improved care and outcomes in IDUs.25 Although IDUs are mostly found in the southern and southwestern provinces and Xinjiang autonomous region in the northwest, former plasma donors (FPDs) are generally rural, poorly educated farmers from the central provinces who were infected through selling plasma. Plasma collectors pooled whole blood, removed the plasma, then returned pooled red blood cells to plasma donors to prevent anemia, allowing more frequent plasma donation. This unhygienic practice infected untold numbers of people in during the 1990s.26,27 The unique circumstances surrounding IDUs and FPDs may have fostered distrust and stigma that interfere with treatment and care, compared to those infected sexually. In addition, gender differences in visit attendance were also observed in our study. Effects of gender on interpretation of illness and health behavior have been documented in other studies.13,28,29 We hypothesize that differences in missed visits reflect differences in social status, access to resources and exertion of control between genders in traditional Chinese culture that allow men to have easier access to information and medical care.30 Finally, in this study, patients above age 60 had a moderately increased risk of missing early visits. This finding is important because older patients account for a rapidly growing proportion of newly diagnosed infections in China.31 Although older patients had been reported to miss fewer visits or have better adherence in studies conducted elsewhere,13,32 China may be different due to the continued social stigma33 and the elderly's dependence on their children to travel and receive medical care.
Other critical factors influencing survival in our study included OI symptoms, baseline CD4 count, and hemoglobin level, which are all markers of disease severity. However, by comparing our results with the risk factors for overall mortality reported by Zhang et al,15 when excluding patients who died within 6 months of ART initiation, the impact of baseline CD4 and OI symptoms was considerably attenuated, and engagement in HIV care emerges as a crucial predictor of survival. The observation that older age was correlated with worse survival is also important in a time when older patients account for an increasing proportion of newly diagnosed HIV cases in China.27
There are some limitations to this study. First, as a retrospective study, we were unable to directly assess reasons for missed visits or address all confounding factors. We have limited information on transportation and distance from patients' homes to clinics, alcohol abuse and mental health, which have been linked with poorer adherence or retention in previous studies.9,21,34 Also, we have mainly focused on patient-level factors, though we recognize that the complex interplay of environmental determinants including social structure, healthcare system and provider factors must also be addressed and should be areas of focus for future studies. Second, a large number of patients from Henan province who started HAART before June 2006 and were predominantly FPDs were excluded from the analysis. Despite this, transmission through plasma donation or blood transfusion still accounts for 46% of the remaining subjects. Considering the association of this mode of transmission with missing early appointments, excluding this group of patients may overestimate the rate of early appointment attendance in our study. In addition, these patients were excluded due to administrative reasons, which is unlikely to impact our evaluation of risk factors of missing visits and mortality. Last, the high proportion of patients with HIV due to plasma donation or blood transfusion may generate concern for the generalizability of these data. However, 8462 patients and 5514 patients who contracted HIV from sexual transmission and drug injection, respectively, were included and compared while controlling for other factors. The association that we found between drug injection, in comparison with sexual transmission, and missing early visits are consistent with the results of other studies.22–24
Understanding the impact of early missed visits on mortality provides insights into areas of focus for the national treatment program in China and other resource-limited settings. Our observations suggest that early missed visits may serve as a proxy for medication adherence and quality of medical care. Risk factors for missed visits will inform future policy and national guidelines to prioritize targeted interventions for this early warning indicator. In addition to intensifying HIV screening and expanding ART coverage, the future of China's ART program rests upon improved appointment adherence to minimize treatment failure and drug resistance while maximizing first-line regimen durability.
The authors are thankful to Ms Christine Korhonen and Dr Xiaoyu Wei at Global AIDS Program, US Centers for Disease Control and Prevention for their careful review of the article, and acknowledge the support of the Doris Duke fellowship through the North Carolina School of Medicine.
1. Zhang F, Dou Z, Ma Y, et al.. Effect of earlier initiation of antiretroviral treatment and increased treatment coverage on HIV-related mortality in China: a national observational cohort study. Lancet Infect Dis. 2011;11:516–524.
2. Giordano TP, White AC Jr, Sajja P, et al.. Factors associated with the use of highly active antiretroviral therapy in patients newly entering care in an urban clinic. J Acquir Immune Defic Syndr. 2003;32(4):399–405.
3. Mugavero MJ, Lin HY, Allison JJ, et al.. Racial disparities in HIV virologic failure: do missed visits matter? J Acquir Immune Defic Syndr. 2009;50:100–108.
4. Giordano TP, Gifford AL, White AC Jr, et al.. Retention in care: a challenge to survival with HIV infection. Clin Infect Dis. 2007;44:1493–1499.
5. Sethi AK, Celentano DD, Gange SJ, et al.. Association between adherence to antiretroviral therapy and human immunodeficiency virus drug resistance. Clin Infect Dis. 2003;37:1112–1118.
6. Valdez H, Lederman MM, Woolley I, et al.. Human immunodeficiency virus 1 protease inhibitors in clinical practice: predictors of virological outcome. Arch Intern Med. 1999;159:1771–1776.
7. Solomon L, Stein M, Flynn C, et al.. Health services use by urban women with or at risk for HIV-1 infection: the HIV Epidemiology Research Study (HERS). J Acquir Immune Defic Syndr Hum Retrovirol. 1998;17:253–261.
8. Cunningham WE, Andersen RM, Katz MH, et al.. The impact of competing subsistence needs and barriers on access to medical care for persons with human immunodeficiency virus receiving care in the United States. Med Care. 1999;37:1270–1281.
9. Giordano TP, Visnegarwala F, White AC Jr, et al.. Patients referred to an urban HIV clinic frequently fail to establish care: factors predicting failure. AIDS Care. 2005;17:773–783.
10. Gardner LI, Metsch LR, Anderson-Mahoney P, et al.. Efficacy of a brief case management intervention to link recently diagnosed HIV-infected persons to care. AIDS. 2005;19:423–431.
11. Brennan AT, Maskew M, Sanne I, et al.. The importance of clinic attendance in the first six months on antiretroviral treatment: a retrospective analysis at a large public sector HIV clinic in South Africa. J Int AIDS Soc. 2010;13:49.
12. Birbeck GL, Chomba E, Kvalsund M, et al.. Antiretroviral adherence in rural Zambia: the first year of treatment availability. Am J Trop Med Hyg. 2009;80:669–674.
13. Cauldbeck MB, O'Connor C, O'Connor MB, et al.. Adherence to anti-retroviral therapy among HIV patients in Bangalore, India. AIDS Res Ther. 2009;6:7.
14. Ma Y, Zhang F, Zhao Y, et al.. Cohort profile: the Chinese national free antiretroviral treatment cohort. Int J Epidemiol. 2011;39:973–979.
15. Zhang F, Dou Z, Ma Y, et al.. Five-year outcomes of the China National Free Antiretroviral Treatment Program. Ann Intern Med. 2009;151:241–251, W-52.
16. Carrieri MP, Raffi F, Lewden C, et al.. Impact of early versus late adherence to highly active antiretroviral therapy on immuno-virological response: a 3-year follow-up study. Antivir Ther. 2003;8:585–594.
17. Ma Y, Zhao D, Yu L, et al.. Predictors of virologic failure in HIV-1-infected adults receiving first-line antiretroviral therapy in 8 provinces in China. Clin Infect Dis. 2010;50:264–271.
18. Lucas GM, Chaisson RE, Moore RD. Highly active antiretroviral therapy in a large urban clinic: risk factors for virologic failure and adverse drug reactions. Ann Intern Med. 1999;131:81–87.
19. Blacher RJ, Muiruri P, Njobvu L, et al.. How late is too late? Timeliness to scheduled visits as an antiretroviral therapy adherence measure in Nairobi, Kenya and Lusaka, Zambia. AIDS Care. 2010;22:1323–1331.
21. Ulett KB, Willig JH, Lin HY, et al.. The therapeutic implications of timely linkage and early retention in HIV care. AIDS Patient Care STDS. 2009;23:41–49.
22. Aloisi MS, Arici C, Balzano R, et al.. Behavioral correlates of adherence to antiretroviral therapy. J Acquir Immune Defic Syndr. 2002;31:S145–S148.
23. Arici C, Ripamonti D, Maggiolo F, et al.. Factors associated with the failure of HIV-positive persons to return for scheduled medical visits. HIV Clin Trials. 2002;3:52–57.
24. Samet JH, Freedberg KA, Savetsky JB, et al.. Discontinuation from HIV medical care: squandering treatment opportunities. J Health Care Poor Underserved. 2003;14:244–255.
25. Wood E, Hogg RS, Lima VD, et al.. Highly active antiretroviral therapy and survival in HIV-infected injection drug users. JAMA. 2008;300:550–554.
26. Dou Z, Chen RY, Wang Z, et al.. HIV-infected former plasma donors in rural Central China: from infection to survival outcomes, 1985–2008. PLoS One. 2010;5:e13737.
27. Lu L, Jia M, Ma Y, et al.. The changing face of HIV in China. Nature. 2008;455:609–611.
28. Mugavero MJ, Lin HY, Allison JJ, et al.. Failure to establish HIV care: characterizing the "no show" phenomenon. Clin Infect Dis. 2007;45:127–130.
29. Kremer H, Sonnenberg-Schwan U, Arendt G, et al.. HIV or HIV-therapy? Causal attributions of symptoms and their impact on treatment decisions among women and men with HIV. Eur J Med Res. 2009;14:139–146.
30. Wang J, Fei Y, Shen H, et al.. Gender difference in knowledge of tuberculosis and associated health-care seeking behaviors: a cross-sectional study in a rural area of China. BMC Public Health. 2008;8:354.
32. Mugavero MJ, Lin HY, Willig JH, et al.. Missed visits and mortality among patients establishing initial outpatient HIV treatment. Clin Infect Dis. 2009;48:248–256.
33. Qian HZ, Wang N, Dong S, et al.. Association of misconceptions about HIV transmission and discriminatory attitudes in rural China. AIDS Care. 2007;19:1283–1287.
34. Geng EH, Bangsberg DR, Musinguzi N, et al.. Understanding reasons for and outcomes of patients lost to follow-up in antiretroviral therapy programs in Africa through a sampling-based approach. J Acquir Immune Defic Syndr. 2010;53:405–411.