Introduction

Patients with human immunodeficiency virus (HIV) cannot currently be free of the virus, so they must regularly visit medical facilities and receive appropriate care to prevent acquired immunodeficiency syndrome (AIDS)-related or non-AIDS-related complications. Patients who do not receive regular HIV care tend to experience adverse outcomes such as antiretroviral therapy failure, hospitalization, AIDS-defining illness, and death [1–3].

One of the approaches to preventing loss to follow-up (LTFU) and facilitating better retention is case management. It is an attempt by health care providers to connect patients to adequate ancillary services. Case management has been shown effective [4–6], but it consists of various complex activities to keep patients from LTFU and the definition of case management differs somewhat among the previous reports. Thus, it is clinically relevant to investigate which element of intervention in case management is particularly associated with these patients’ engagement in care.

Meanwhile, patients with HIV and poor mental health or psychological distress are prone to be lost to follow-up [7–9] and Japanese local governments staff therapists to provide psychotherapy at no charge as a mental health service upon requests from HIV care medical staff members. Therefore, case managers at HIV care facilities can easily utilize mental health services as resources, despite little evidence that mental health service introduction improves patient retention. Our aim is to investigate the association between mental health service implementation for patients by case managers and LTFU in HIV care.

Methods

Results

Of the 1715 patients with HIV in the hospital cohort, 1620 were eligible for the present study. Their mean age was 37.1 years and 1545 (95.4%) were male. The route of HIV infection was homosexual transmission in 1238 (76.4%) of patients, heterosexual transmission in 277 (17.1%), blood products in 34 (2.1%), and unidentified in 71 (4.4%).

A total of 88 LTFU cases were identified and 264 controls were selected from the cohort. The incidence of LTFU was calculated to be 1.32 per 100 person-years. The median time to LTFU was 14 months (interquartile range, 1.6-33 months). Forty LFTU cases (45.5%) resulted in absence within 1 year of the first visit. Six patients had only one hospital visit. The patients’ demographics, disease status, and treatment characteristics are summarized in Table 1. Patients in the LTFU cases had an average age of 32.7 years (standard deviation, 9.3 years) and included 86 males (97.7%). Patients in the LTFU cases were younger than the controls and had a larger proportion of patients who underwent voluntary HIV testing or who lacked regular employment. Patients in the LTFU cases also had higher CD4+ lymphocyte counts and were less likely to have an AIDS diagnosis or to have received antiretroviral therapy than the controls. Sixty-eight (91.9%) of the 74 LTFU case patients without antiretroviral therapy had CD4+ lymphocyte counts higher than the threshold for the therapy recommended by the guidelines at the time LTFU occurred or were within 90 days from their first visit. Fourteen LTFU case patients had been on anti-retroviral therapy for the median of 31 months (range, 2-80 months).

Independent variables for LTFU are displayed in Table 2. Stepwise multivariate-adjusted conditional logistic regression revealed that the risk of LTFU in patients who had received mental health services implemented by a case manager was significantly low (0.35 [0.16-0.76], p = 0.008). Loss to follow-up occurred more frequently in patients who did not receive antiretroviral therapy (adjusted OR [95% confidence interval], 7.51 [3.34-16.9], p < 0.001), who were younger than 30 years old (2.74 [1.36-5.50], p = 0.005), and who lacked regular employment (3.38 [1.58-7.23], p = 0.002).

Discussion

This study found that patients who utilized mental health services introduced by a case manager were less likely to be LTFU.

Previous studies have demonstrated that case management is associated with better retention and survival of patients with HIV [11,12]. However, case management has been differently operationalized and which specific intervention by case managers improve patients’ prognosis has not been fully elucidated [13,14]. What our result adds to this topic is that the mental health services based on a case manager’s assessment of each patient’s psychological aspects increased the chance for patients to receive regular care.

There could be several reasons for good retention of the patients who receive mental health services. Improved mental health as the result of the services could be related to good retention [15–17] and implementation of the services per se would be responsible for good engagement in care through making patients realize their own problems and tackling them. An understanding of a patient’s psychological aspects by the case manager and other medical staff members through communication with therapists would also contribute to better continuity of care. We cannot say that mental health service implementation directly reduced LTFU because of unmeasured confounding factors such as a patient’s attitude to accepting a case manager’s advice. However, our results suggested that patients who did not accept mental health services were more at-risk for LTFU.

Patients with HIV younger than 30 years of age or without regular employment were less likely to engage in regular hospital visits. These results agree with previous reports that showed the independent factors for LTFU were young age and low socioeconomic status [18–22]. Similar to previous reports [21,22], not taking antiretroviral therapy also strongly predicted LTFU. Patients with HIV who did not receive antiretroviral therapy were over seven times more likely to withdraw from regular care. They were speculated to be asymptomatic because, in our data, 93.2% did not have an AIDS diagnosis. Furthermore, because these patients were free from the side effects of antiretroviral drugs, they might not have felt an urgent need for strict regular checkups. We cannot determine the exact reason why some patients were not receiving antiretroviral therapy at the time of LTFU, because whether or not to start the therapy depends partially on individual preference. However, they were not assumed to be eligible or prepared for the therapy because most of them had CD4+ lymphocyte counts higher than the threshold for the therapy recommended by the guidelines of the time [23], or were within 90 days from their first visit.

This study has some limitations. First, because it is not a randomized-controlled study but an observational study, we cannot determine the causal relationship between mental health service induction and good retention. However, it is obviously ineffective to deliver patient-centered psychotherapy to patients without the need and it is unethical to withhold it from patients who need it. An observational study is reasonable to investigate the association between the implementation of psychotherapy and retention in actual clinical practice. Second, the survival of the patients was not investigated and we could not obtain the prognosis of LTFU cases. However, the growing number of studies including ones without complete follow-up systems, such as nationwide databases, indicates that LTFU is associated with worse survival, which can support the clinical validity of the present findings [24–27]. The third limitation is generalizability. Our findings would be applicable to HIV care facilities where case manager have good access to mental health services, but the availability, contents, and charge of mental health services differ among settings and influence results.

Our study has several strengths. All of our LTFU cases and controls received case management, unlike most previous studies that evaluated the effectiveness of case management. Consequently, we were able to demonstrate an association between retention and mental health service use as an element of case management more clearly. Therapeutic trends such as criteria for antiretroviral therapy initiation, pill burden, and drug side effects are always changing and they must have affected results; however, this bias was accounted for because we matched entry dates into the cohort and time spent in the cohort between LTFU cases and controls. We suspected that CD4+ lymphocyte count and antiretroviral therapy at cohort entry affect the subsequent continuity of care and we dealt with these factors as time-dependent variables. Thus, we could evaluate the influence of these factors at the time LTFU occurred more appropriately.

Conclusions

Treatment of HIV has improved drastically, but care cannot be optimally offered if patients do not regularly keep hospital appointments. Paying attention to patients’ psychological problems and implementing adequate resources such as mental health services have significance in case management for better retention, and healthcare professionals must recognize that patients who do not utilize mental health services are at high risk of LTFU.

Acknowledgments

The authors thank Y. Shimoji, M. Higashi, and T. Harukawa for their data collection assistance.