Associations between perceived barriers and benefits of using HIV pre-exposure prophylaxis and medication adherence among men who have sex with men in Western China

The report of the Joint United Nations Program on HIV/AIDS [1] points out that the number of HIV infections reached 1.8 million in 2016 and men who have sex with men (MSM) have a higher risk of infection. The Progress Report on AIDS Prevention and Treatment in China [2] shows that the HIV epidemic currently maintains a low- prevalence trend, with a higher-prevalence among MSM, especially in the western part of China [3]. In 2006, MSM accounted for 2.5% of new HIV infections in China, and the rate increased dramatically to 25.5% in 2017 [4, 5].

Pre-exposure prophylaxis (PrEP) has been considered the most promising biomedical HIV prevention strategy so far. The antiviral drug tenofovir (TDF) used in PrEP and combined therapy with TDF and emtricitabine (TDF-FTC) have been approved by the USFDA as preventive drugs for the MSM population [6], and in 2014, comprehensive PrEP guidelines were released [7]. Many international clinical studies have shown the effectiveness of PrEP in different groups, including MSM [8‐10], women [11] and injection drug users [12, 13].

In addition, studies have also suggested that adherence to medication is crucial for the efficacy of PrEP [9, 11, 14‐16]. The iPrEP study demonstrated that the once-daily oral PrEP drug resulted a 44% reduction in the incidence of HIV among MSM and transgender women, while pill use on 90% or more of days showed greater protection (73%), and it increased to 92% efficacy for participants with detectable drug levels [10]. However, a study conducted among African women unable to show the efficacy of oral PrEP due to their low adherence levels [17]. Therefore, an investigation into barriers and facilities associated with adherence to PrEP is required.

Factors from the perspectives of social, behavioural and psychological areas associated with PrEP adherence have been identified in clinical trials [9, 18‐20]. However, in China, only one open-label clinical trial has been conducted among HIV-uninfected MSM in western China to assess the efficacy and safety of oral TDF in preventing HIV infection [21], but little is known about the barriers and facilities of PrEP adherence in China. Based on the fact that PrEP clinical trials have not been widely conducted in China, assessing participants’ feelings and attitudes during medication among the very first group of people involved in taking PrEP drugs is important and necessary. For instance, participants who perceive that there are side effects of drugs tend to show low adherence to medication, while sound doctor-patient relationships are a facilitator of adherence [22, 23].

Therefore, in this study, we aimed to investigate the changes in barriers and benefits that HIV-uninfected MSM perceive during the course of medication and examine the associations between perceived barriers and benefits of oral PrEP and actual medication adherence to provide a basis and some guidance for application and promotion of PrEP strategies among MSM in China.

Of 575 participants who completed the baseline survey in the TDF group, 141 who did not participate in at least one follow-up and 23 who did not report the number of missed doses at least once were excluded. Thus, a total of 411 MSM involved in at least one follow-up with a maximum of 8 and an average of 3.54 were included in the final analysis. A total of 1561 follow-up observation points were provided.

The demographical and HIV-related characteristics are shown in Table 1. The average age of the 411 MSM was 29.34 ± 7.85 years, and their age ranged from 18 to 61 years. Among the participants, 73.90% were urban residents, and 26.10% were rural residents; 90% graduated from high school and above, and 78.83% were unmarried. Nearly half (48.27%) had an average monthly income above 3000 yuan. The scores of HIV knowledge ranged from 0 to 13 with a median of 9, and only 26.76% had scores above 11. While 64.63 and 81.91% of participants declared that they had ever engaged in HIV counselling and testing, respectively, only 56.97% believed that AIDS is a great threat to themselves and their families. Only 40.63% had only one male sexual partner in the past six months, but only 9.34% had been diagnosed with an STD in the past six months.

The scores of the barriers and benefits participants perceived at baseline and during the follow-up are shown in Table 2. The changes in the scores of “worries about drug side effects” and “positive expectation about the efficacy of the drugs” over the follow-up are shown in Fig. 1. At baseline, when participants had not yet started to take PrEP, the score for “worries about drug side effects” was 2.97 ± 1.15, but it showed a decreasing trend during the follow-up. Participants had a relatively high expectation about the efficacy of the drugs at baseline (3.17 ± 1.36), and even though the score was 2.89 ± 1.41 at first follow-up, there was an increasing trend throughout the follow-up, which reached 3.72 ± 1.37 at the end. The scores of “disliked taking drugs” and “burden perceived during medication” fluctuated at a low and moderate level throughout the follow-ups, respectively. The score of “relationships between doctor and participants” was high both at baseline (4.20 ± 1.13) and throughout the entire follow-up.

The average medication rate for the entire cohort was 0.62 ± 0.37, and its changes during the follow-ups are shown in Table 2. The results of the linear mixed model used to analyse the factors associated with medication adherence are shown in Table 3. The changes of predictive medication rate over the follow-up are shown in Fig. 1. The β coefficient for “time” (meaning follow-up times and coded as 1 to 8) was significant (β = 0.014; 95% CI = 0.005–0.024), indicating that the medication rate increased slightly during the follow-up. After adjusting for potential confounding variables, participants who were divorced showed a higher medication rate (compared with those who were unmarried, β = 0.179; 95% CI = 0.047–0.311). “Positive expectations towards medication”, as a time-varying value, was positively associated with the medication rate over time (β = 0.033; 95% CI = 0.020–0.047), indicating that those with higher scores on positive expectations as to the efficacy of the medicine over time reported better adherence to medication over time. In addition, the more worried the participant was about drug side effects during the follow-up, the lower the medication rate (β = − 0.019; 95% CI = − 0.035−− 0.003). However, the dislike of taking drugs, burden perceived during medication, and the relationship between the doctor and participant showed no correlation with the medication rate (P > 0.05).

This study is among the first to examine the potential factors related to PrEP adherence among MSM in a clinical trial in China, and its findings suggest that MSM’s positive expectations towards medication was a facilitator, while their worries about drug side effects was a barrier to PrEP adherence during the whole course of medication.

In the present study, a positive expectation towards medication refers to their trust in the efficacy of the drugs, which was evaluated only at a moderate level during the first follow-up but increased as the follow-up continued. The medication rate became higher along with an increase in their expectations during the follow-up, suggesting that it is still urgent to make efforts to publicize the effectiveness of PrEP drugs. Although many clinical trials of PrEP have been conducted worldwide and TDF was first approved by the FDA as a preventive drug for the MSM population in 2012, this is the first clinical research on PrEP conducted in China and according to the previous surveys on the willingness to use PrEP among MSM in China, only 22% of MSM (308/1402) reported that they had heard of PrEP [24], and a similar rate was observed among female sex workers (16.5%, 264/1611) [25] indicating that MSM in China only have a relatively low awareness of PrEP and that the majority of them have very little understanding of this new biomedical HIV prevention strategy. Therefore, providing more clinical evidence to them may contribute to improving their adherence to medication. It has always been said that the efficacy of drugs depends on patient adherence, but efficacy and adherence may complement each other since positive expectations as to their efficacy encourages people to be more likely to take the medicine on their own initiative.

In addition, the more worried patients are about side effects, the lower the medication rate, which is consistent with not only those studies examining factors associated with willingness to use PrEP [26] but also results found in clinical trials [27, 28]. Unlike therapeutic drugs, few people would take a preventative drug if it has serious side effects, which is in line with the findings of Mustanski et al. [29] that PrEP is more likely to be used as a preventative strategy for HIV if it has a low burden of side effects and satisfactory benefits are perceived. The worries about drug side effects showed a declining trend in the MSM during the follow-ups, which may be related to the low incidence of adverse events during medication. In our entire follow-up, the main adverse events were nausea, vomiting and diarrhoea, all with a low incidence, and there were no serious adverse events reported. Therefore, in addition to positive expectations of drug efficacy affecting the medication rate, their confidence in the safety of the drugs was also an important facilitator of their medication, which suggested that we should provide them correct information about this drug, inform them of possible side effects, and help them during the follow-ups, thereby reducing their worries about the efficacy and safety of the drugs and improve their adherence to medication.

There was no relationship between disliking taking drugs and the medication rate. In our study, the dislike of taking drugs referred to disliking the drug formulation, taste and pill size. The TDF used in this study were white tablets with 300 mg/tablet. During the follow-ups, the score of the dislike of taking drugs was lowest, which means the drug formulation, taste and size were acceptable by participants to some extent.

No association between the burden perceived during medication and the medication rate was observed. We hypothesized that taking medicine would give their partners a sense that they are in HIV at-risk status, which may impact their adherence. However, we did not find any such relationship and considered that this may be related to the dosing strategy, in which daily oral drugs had more flexibility when compared with dosing ‘on demand’, another dosing strategy recommended for MSM (a double dose of TDF/FTC 2–24 h before each sexual intercourse, followed by two single doses of TDF/FTC, 24 and 48 h after the first drug intake), and with this dosing strategy (dosed ‘on demand’), Mutua et al. [30] found that the burden perceived during medication reduced adherence.

However, the average medication rate during the entire follow-up was not very high. We believed that, other than the factors that we have examined in this study, there were some other barriers that existed and impacted the medication rate. In this study, we mainly focused on some attitudinal factors during medication and examined their relationships with adherence, which could provide some guidelines as to the implementation of clinical trials in China, and thus, it is the very first step in the promotion of PrEP in China. Nevertheless, medication is a long-term process, and long-term adherence to medication may be affected by many factors. For instance, patients may forgot to take their medicine, which has been reported in some studies [28, 31] and in our study; when we asked participants why they missed their doses, some reported that they forgot to take them sometimes. Thus, in the future, interventions should be developed to help remind people to take their medicine. In addition, supplying patients with additional correct information about the benefits and potential risks of the drugs, provided to both MSM and clinicians, their confidence in the drug efficacy could increase, and at the same time, their adherence could be improved.

How to obtain and maintain high medication adherence is the key in PrEP to achieve an effective reduction in HIV infection. The results of our study showed that the overall medication rate in the daily medication group is not high. Worries about the efficacy and safety of PrEP were factors influencing the medication rate. For MSM in China, we should work to increase their understanding of the effectiveness and safety of PrEP and increase their confidence in PrEP, thereby improving their medication adherence.