https://orcid.org/0000-0003-3036-5740
Figures
Abstract
Objective
To measure adherence self-efficacy (ASE) to antiretroviral therapies (ART) and evaluate the relationship between ASE, depression, perceived social support, and HIV viral suppression among adolescents and young adults living with HIV (ALHIV) in Kenya.
Methods
ALHIV were recruited from nine health facilities in western Kenya. Participants completed behavioral surveys at enrollment, and study team members extracted viral load data from a national database. ASE was assessed using a modified HIV-Adherence Self-efficacy Assessment Survey (HIV-ASES), and depression was assessed as a score of ≥ 10 using the PHQ-9. Linear mixed effects regression modeling and general linear mixed effects regression modeling, clustering by facility, were used to determine associations between ASE scores, viral suppression, and correlates of interest.
Results
Overall, 987 ALHIV age 15 and older were included in this study, 70% were female, 58% ages 15−19, and 57% were attending or had completed at least secondary school. Ninety-six percent had ASE data, 73% (517/703) were virally suppressed, 90% (888/987) scored 9 or below on the PHQ-9, 47% (460/987) reported high perceived social support, and 65% (645/987) were classified as having orphan status. ALHIV who had moderate-to-severe depression had a mean ASE score that was 13.41 points lower (95% CI: −20.12 – −6.52, p < 0.001) than those with none-or-mild depression. Female ALHIV had higher odds of viral suppression (adjusted OR: 1.55, 95%CI: 1.07–2.25, p = 0.02) as did ALHIV with higher social support (adjusted OR: 1.67, 95%CI: 1.17–2.40, p = 0.005). There was no significant association identified between ASE and viral suppression.
Citation: Oja D, Jiang W, Richardson BA, Badia J, Kibugi J, Beima-Sofie K, et al. (2026) Psychosocial correlates of adherence self-efficacy and HIV viral suppression among adolescents and young adults in Western Kenya. PLoS One 21(2): e0341269. https://doi.org/10.1371/journal.pone.0341269
Editor: Makandwe Nyirenda, South African Medical Research Council, SOUTH AFRICA
Received: September 29, 2025; Accepted: January 5, 2026; Published: February 3, 2026
Copyright: © 2026 Oja et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data that support the findings of this study are available upon reasonable request from The Global Center for Integrated Health of Women, Adolescents, and Children (Global WACh) at the University of Washington (email: gwach@uw.edu). Data cannot be made publicly available at this time due to the risk of participant identification and planned future analysis.
Funding: The DiSC study was supported by funding from NICHD (UG3 HD096906/HD/NICHD NIH HHS/United States). First author DO is supported by National Institute Of Nursing Research of the National Institutes of Health under Award Number T32NR020776. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Adherence to antiretroviral treatment (ART) is critical in maintaining HIV viral suppression and improving health outcomes for adolescents and young adults living with HIV (ALHIV). Low ART adherence increases viral replication and disease progression, reducing quality of life and increasing mortality among individuals living with HIV [1]. Multiple studies report that ART adherence among ALHIV is significantly lower than adults [2–4], with barriers to adherence across personal, caregiver and health system-related domains, including stigma, forgetfulness, secrecy, ART side effects, inconvenience, and lack of support [3]. Targeting self-efficacy could be a focal point in addressing some of these ART adherence barriers at the individual level while enabling ALHIV to self-manage their HIV.
Self-efficacy is defined as an individual’s belief in their ability to implement a specific behavior or a set of behaviors [4]. Among adolescents and young adults, self-efficacy is related to positive outcomes such as improved academic performance, emotional health, and quality of life [5]. Self-efficacy score measures have been used to predict outcomes of various chronic conditions such as asthma, type 2 diabetes, hypertension, and other chronic health conditions [6–9].
Within the broader construct of self-efficacy, researchers have examined specific concepts of adherence self-efficacy (ASE) relating to treatment plans. ASE is defined as “confidence in one’s ability to adhere to treatment plans” and has been shown to be an important predictor of medication adherence for HIV [10,11]. Among adults living with HIV, ASE is a strong predictor of initiation and maintenance of ART [12,13]; however, there have been few studies of ASE among adolescents on ART in Kenya [14].
Interventions to target ASE as a mechanism to improve ART adherence may be a potential avenue to increase the self-management of HIV for adolescents and youths, however contributing factors of mental health and social support may attenuate these relationships. Thus, in this paper, we measured ASE to ART and evaluated the relationship between ASE, depression, perceived social support, and HIV viral suppression among ALHIV in Kenya.
Materials and methods
Study population and design
This study utilized baseline data from Kenyan ALHIV enrolled in the Data-informed Stepped Care (DiSC) to Improve Adolescent HIV Outcomes cohort study [15]. ALHIV 10–24 years of age were recruited from nine HIV care and treatment facilities located in Kisumu, Homabay, and Migori counties, from April 1, 2019, to February 28, 2020. For this analysis, participants aged 15 years and older were included, consistent with prior use and validation of the ASE scale. Data for this analysis were accessed on October 20, 2022, from the DiSC study data manager at the University of Washington.
Conceptual framework
The analysis is guided by CK Ewart’s Social Action Theory (SAT), which emphasizes a model of behavior change and the social context and support that assist in achieving and maintaining that behavior change [16–18]. According to SAT, health behaviors result from the interaction of three domains: contextual influences, the process of self-change, and self-regulation [18]. The contextual influences encompass the context in which the health behavior occurs, such as individuals’ background and demographics, life stressors, and mental health state. The process of self-change includes self-change processes such as self-efficacy and social support. Self-regulation is an action state, which includes actions such as adherence to ART to reduce HIV viral load. Thus, intended health behaviors are shaped by the social-environmental system and interpersonal factors, which can both hinder or facilitate a behavior change [16]. SAT has been utilized in various studies to promote healthy behaviors such as ART adherence [19,20]. Fig 1 represents measures assessed in this study based on SAT.
Data collection
Behavioral surveys were administered by trained study staff via tablets at enrollment. Viral load (VL) data were obtained from the National AIDS and STI Control Programme database. VL collected within three months before or after completion of the DiSC baseline survey was used for the analysis. If there were two VLs collected within a three-month window, the VL collected closest to the survey date was used.
Measures
The outcomes in this study were ASE score, measured by the HIV-Adherence Self-efficacy Assessment Survey (HIV-ASES) [21], and VL suppression. HIV-ASES is a psychometric scale that has been previously validated to assess people living with HIV’s confidence to carry out important behaviors around adherence to treatment plans [14,21,22]. In this study of adolescents, we omitted one of the twelve ASES questions, “How confident have you been that you get something positive out of your participation in treatment, even if the medication you are taking does not improve your health?” citing concern that the question would be interpreted as stating ART does not improve health. We categorized participants as having high ASE if the HIV-ASES total score was above 90 [14]. VL suppression was defined as having < 200 copies of HIV RNA/ ml of blood based on Kenya’s HIV Treatment Guideline definition [23].
Predictors analyzed in this study were age groups, sex, education level, orphan status, depression (using the Patient Health Questionnaire-9 (PHQ-9) [24]), and social support (using the Multidimensional Scale of Perceived Social Support (MSPSS) [25]). Orphan status was defined based on the participant’s report of having lost one or both parents. In alignment with clinical guidelines, a PHQ-9 score of 9 or below was considered as having none to mild depressive symptoms, and a score of 10 or higher was considered as moderate to severe depressive symptoms [24]. The MSPSS is a brief questionnaire designed to measure perceptions of social support from three sources: family, friends, and significant others. In this study, high perceived social support was categorized as those with a score above median while low perceived social support were classified as those with a score below median.
Statistical analysis
Cumulative ASE scores were described in medians with interquartile ranges (IQR) and proportions of viral suppression were described in counts and proportions. Correlates were identified using linear mixed effects regression models and general linear mixed effects regression models, clustered by facility. Variables with p-values <0.10 in the bivariate analyses were included in multivariable models for ASE and viral suppression. All analyses were done in R [26].
ASE scores were missing for 4% of participants. We compared participants with and without ASE data across key predictors including age, gender, depression and perceived social support score. Preliminary analyses showed no significant differences in key predictors between participants missing ASE data vs. without, suggesting that the ASE data were likely missing completely at random.
Ethical approval
We obtained approvals from the University of Washington Institutional Review Board (STUDY00005767) and Maseno University Ethics Review Committee (ERC) (MSU/DRPI/MUERC/00799/19). Caregivers and adolescents provided informed consent and assent until August 2019. On August 22, 2019, we received approval from Maseno ERC to waive parental consent of adolescents ages 15–17 who came to the clinic without a caregiver.
Results
Participant characteristics
Survey data from 987 ALHIV ages 15 and older were included in this analysis. The median age of participants was 18 years (IQR:16–21) (Table 1). Most participants (70%) were female. Five hundred and sixty-four participants (57%) currently attended or had completed secondary school. Approximately 90% reported no or mild depressive symptoms over the past two weeks. Most participants (65%) reported that one or both parents were deceased. ASE cumulative scores were relatively high, with a median score of 99 out of 110 (IQR: 90–105). The median MSPSS score was 43 out of 60 (IQR: 38–48).
Itemization of the HIV adherence self-efficacy scale
When we evaluated each of the 11 HIV-ASES scale questions in an itemized manner, we found that for all questions, except one, the median score was 10. Question three assessed incorporating treatment in daily routine, including in front of people who don’t know their HIV status, and the median score was 5 (IQR: 0–10) (Table 2).
Correlates of HIV adherence self–efficacy
We evaluated the association between the cumulative ASE score and risk factors, including age groups, sex, education levels, depressive symptoms, orphan status, and perceived social support (Table 3). In linear mixed effects models clustered by facility, ALHIV who had moderate-to-severe depression had a mean ASE score that was 13.41 points lower (95% CI: – 20.12 – −6.52, p < 0.001) than those with none-or-mild depression. There were no statistically significant differences in cumulative ASE scores by age group, sex, education, orphan status, and those with higher than median and lower than median levels of perceived social support in a multivariate analysis.
Correlates of viral suppression
A total of 703 (71%) ALHIV had viral load data within 3 months of baseline surveys, and 74% (n = 517) were virally suppressed. ALHIV with higher than median social support score had higher odds of viral suppression compared to those with lower than median social support score (78% vs. 68% respectively, adjusted OR [aOR]: 1.67, 95%CI:1.17–2.49, p = 0.009) (Table 4). Additionally, female ALHIV had higher odds of viral suppression (76%) than male ALHIV (68%) (aOR:1.55, 95%CI: 1.07–2.25, p = 0.02). Age groups, education levels, severity of depression, and ASE scores were not associated with viral suppression in univariate regression models.
Discussion
This cross-sectional analysis of a large cohort of ALHIV found a positive correlation between adherence self-efficacy and mental health and an association between social support and viral suppression. Our findings indicate relatively high levels of adherence self-efficacy to ART among ALHIV, with a single item driving differences in ASE scores: incorporating treatment into a daily routine even if it meant taking medication in front of others who were unaware of their HIV status. Similar findings of difficulty taking medication in public spaces, using the HIV-ASES psychometric tool, have been reported among ALHIV in Nairobi [27,28]. Perceived stigma and fear of unintended disclosure associated with taking ART in public have been noted among ALHIV globally as contributors to ART adherence [29–31]. Well-designed interventions that target internal stigma reduction and enable adolescents to take their ART in whatever spaces they choose are imperative to improve HIV outcomes among ALHIV.
Our findings indicate that depression is associated with reduced ASE scores. Consistent with our findings, other studies have also observed that depression is linked to lower levels of ART adherence among individuals living with HIV [32–34], and further, self-efficacy has been found to mediate the association between depression and treatment adherence in chronic illnesses such as heart failure and hypertension [11,35]. Identifying and treating mental health conditions such as depression potentially impacts self-efficacy skills, which leads to increased medication adherence and, ultimately, improved HIV outcomes among ALHIV. Thus, at the health systems level, structural interventions, such as prioritization of treatment for mental health conditions, which target emotional and physiological states, could be utilized to enhance self-efficacy for HIV management.
Social support has similarly been noted to have a robust effect on health, with its role as a protective factor against morbidity and all-cause mortality [36,37]. SAT and Social Cognitive Theory both indicate that social support likely influences self-efficacy beliefs [18,38,39]. However, we did not find that the perceived social support score was associated with the adherence self-efficacy score, though other studies have found that social support had a positive relationship with treatment self-efficacy [39,40]. Similar to other studies, we found that ALHIV with high perceived social support had higher odds of suppressed viral loads [41,42]. This suggests that social support plays a critical role in health outcomes, potentially through increased motivation, emotional well-being, or practical support that facilitates treatment adherence. Thus, these findings underscore the importance of a strong support network for ALHIV to achieve viral suppression and improve health outcomes.
Our study had generally high self-efficacy scores and viral suppression and may not have had sufficient variability or statistical power to detect the impact of self-efficacy on viral suppression. Other studies have shown that ART adherence self-efficacy has been associated with increased self-reported adherence and VL suppression among ALHIV in Kenya and among adults in the United States [14,43]. Studies of adolescents and youth in Thailand report that adherence self-efficacy is associated with increases in overall quality of life and physical health [28]. Given self-efficacy’s mediating role on SAT and based on literature findings as discussed above, it is a potential target area to increase self-management of HIV for adolescents and youth.
This study has strengths and limitations. This analysis was a cross-sectional study thus, temporal changes in variables were not assessed. Participants of our study were adolescents and youth in western Kenya and may not be generalizable to the broader population. Further, participants filled out a lengthy survey that consisted of multiple measures and could have been impacted by participant fatigue. ALHIV were also not required to complete all the survey questions; thus, those with missing data may have different experiences or opinions.
Conclusion
Emotional well-being and social support are essential to improve adherence, self-efficacy, and viral suppression among ALHIV. To meet viral suppression goals in ALHIV, interventions targeting perceived stigma reduction and increasing self-efficacy may empower and enable adolescents and youths to take their HIV medication in any space they desire.
References
- 1. Nabunya P, Bahar OS, Chen B, Dvalishvili D, Damulira C, Ssewamala FM. The role of family factors in antiretroviral therapy (ART) adherence self-efficacy among HIV-infected adolescents in southern Uganda. BMC Public Health. 2020;20(1):340. pmid:32183762
- 2. Zurbachew Y, Hiko D, Bacha G, Merga H. Adolescent’s and youth’s adherence to antiretroviral therapy for better treatment outcome and its determinants: multi-center study in public health facilities. AIDS Res Ther. 2023;20(1):91. pmid:38115098
- 3. Ammon N, Mason S, Corkery JM. Factors impacting antiretroviral therapy adherence among human immunodeficiency virus-positive adolescents in Sub-Saharan Africa: a systematic review. Public Health. 2018;157:20–31. pmid:29501984
- 4. Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev. 1977;84(2):191–215. pmid:847061
- 5. Efthymiou V, Charmandari E, Vlachakis D, Tsitsika A, Pałasz A, Chrousos G, et al. Adolescent Self-Efficacy for Diet and Exercise Following a School-Based Multicomponent Lifestyle Intervention. Nutrients. 2021;14(1):97. pmid:35010972
- 6. Campbell TS, Lavoie KL, Bacon SL, Scharf D, Aboussafy D, Ditto B. Asthma self-efficacy, high frequency heart rate variability, and airflow obstruction during negative affect in daily life. Int J Psychophysiol. 2006;62(1):109–14. pmid:16632007
- 7. Huang J, Ding S, Xiong S, Liu Z. Medication Adherence and Associated Factors in Patients With Type 2 Diabetes: A Structural Equation Model. Front Public Health. 2021;9:730845. pmid:34805063
- 8. Warren-Findlow J, Seymour RB, Brunner Huber LR. The association between self-efficacy and hypertension self-care activities among African American adults. J Community Health. 2012;37(1):15–24.
- 9. Kara S. General self-efficacy and hypertension treatment adherence in Algerian private clinical settings. J Public Health Afr. 2022;13(3):2121. pmid:36277949
- 10. Holley S, Knibb R, Latter S, Liossi C, Mitchell F, Radley R, et al. Development and validation of the Adolescent Asthma Self-Efficacy Questionnaire (AASEQ). Eur Respir J. 2019;54(1):1801375. pmid:31048348
- 11. Son Y-J, Won MH. Depression and medication adherence among older Korean patients with hypertension: Mediating role of self-efficacy. Int J Nurs Pract. 2017;23(3):10.1111/ijn.12525. pmid:28194846
- 12. Náfrádi L, Nakamoto K, Schulz PJ. Is patient empowerment the key to promote adherence? A systematic review of the relationship between self-efficacy, health locus of control and medication adherence. PLoS One. 2017;12(10):e0186458. pmid:29040335
- 13. Langebeek N, Gisolf EH, Reiss P, Vervoort SC, Hafsteinsdóttir TB, Richter C, et al. Predictors and correlates of adherence to combination antiretroviral therapy (ART) for chronic HIV infection: a meta-analysis. BMC Med. 2014;12:142. pmid:25145556
- 14. Gitahi-Kamau N, Wahome S, Memiah P, Bukusi EA. The Role of Self-Efficacy in HIV treatment Adherence and its interaction with psychosocial factors among HIV Positive Adolescents in Transition to Adult Care in Kenya. Vulnerable Child Youth Stud. 2022;17(4):308–19. pmid:36439942
- 15. Wilson K, Agot K, Dyer J, Badia J, Kibugi J, Bosire R, et al. Development and validation of a prediction tool to support engagement in HIV care among young people ages 10-24 years in Kenya. PLoS One. 2023;18(6):e0286240. pmid:37390119
- 16. Mukumbang FC, Van Belle S, Marchal B, van Wyk B. Exploring “generative mechanisms” of the antiretroviral adherence club intervention using the realist approach: a scoping review of research-based antiretroviral treatment adherence theories. BMC Public Health. 2017;17(1):385. pmid:28472938
- 17. Remien RH, Stirratt MJ, Dognin J, Day E, El-Bassel N, Warne P. Moving from theory to research to practice. Implementing an effective dyadic intervention to improve antiretroviral adherence for clinic patients. J Acquir Immune Defic Syndr. 2006;43(Suppl 1):S69–78.
- 18. Ewart CK. Social action theory for a public health psychology. Am Psychol. 1991;46(9):931–46. pmid:1958012
- 19. Johnson MO, Catz SL, Remien RH, Rotheram-Borus MJ, Morin SF, Charlebois E, et al. Theory-guided, empirically supported avenues for intervention on HIV medication nonadherence: findings from the Healthy Living Project. AIDS Patient Care STDS. 2003;17(12):645–56. pmid:14746658
- 20. Umar E, Levy JA, Bailey RC, Donenberg G, Hershow RC, Mackesy-Amiti ME. Virological Non-suppression and Its Correlates Among Adolescents and Young People Living with HIV in Southern Malawi. AIDS Behav. 2019;23(2):513–22. pmid:30132172
- 21. Johnson MO, Neilands TB, Dilworth SE, Morin SF, Remien RH, Chesney MA. The role of self-efficacy in HIV treatment adherence: validation of the HIV Treatment Adherence Self-Efficacy Scale (HIV-ASES). J Behav Med. 2007;30(5):359–70. pmid:17588200
- 22. Ashaba S, Baguma C, Tushemereirwe P, Nansera D, Maling S, Zanon BC, et al. Correlates of HIV treatment adherence self-efficacy among adolescents and young adults living with HIV in southwestern Uganda. PLOS Glob Public Health. 2024;4(9):e0003600. pmid:39231105
- 23.
Ministry of Health Kenya. Kenya HIV Prevention and Treatment Guidelines. 2022.
- 24. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13. pmid:11556941
- 25. Zimet GD, Powell SS, Farley GK, Werkman S, Berkoff KA. Psychometric characteristics of the Multidimensional Scale of Perceived Social Support. J Pers Assess. 1990;55(3–4):610–7. pmid:2280326
- 26.
Team RC. R: A language and environment for statistical computing. R Foundation for Statistical Computing; 2013.
- 27. Gitahi-Kamau N, Wahome S, Bukusi EA, Ngure K. Determinants of antiretroviral therapy adherence among older adolescents living with HIV in Kenya during the transition to adult care; An observational study. J AIDS HIV Res. 2020;12(2):24–33. pmid:34540322
- 28. Aurpibul L, Detsakunathiwatchara C, Khampun R, Wongnum N, Chotecharoentanan T, Sudjaritruk T. Quality of life and HIV adherence self-efficacy in adolescents and young adults living with perinatal HIV in Chiang Mai, Thailand. AIDS Care. 2023;35(3):406–10. pmid:35567281
- 29. Madiba S, Josiah U. Perceived Stigma and Fear of Unintended Disclosure are Barriers in Medication Adherence in Adolescents with Perinatal HIV in Botswana: A Qualitative Study. Biomed Res Int. 2019;2019:9623159. pmid:31886271
- 30. Apondi E, Wachira J, Ayikukwei R, Kafu C, Onyango J, Omollo M, et al. Barriers to ART adherence among school students living with HIV in Kenya. Afr J AIDS Res. 2021;20(3):232–7. pmid:34635018
- 31. Robinson A, Cooney A, Fassbender C, McGovern DP. Examining the Relationship Between HIV-Related Stigma and the Health and Wellbeing of Children and Adolescents Living with HIV: A Systematic Review. AIDS Behav. 2023;27(9):3133–49. pmid:36917426
- 32. Shubber Z, Mills EJ, Nachega JB, Vreeman R, Freitas M, Bock P, et al. Patient-Reported Barriers to Adherence to Antiretroviral Therapy: A Systematic Review and Meta-Analysis. PLoS Med. 2016;13(11):e1002183. pmid:27898679
- 33. Uthman OA, Magidson JF, Safren SA, Nachega JB. Depression and adherence to antiretroviral therapy in low-, middle- and high-income countries: a systematic review and meta-analysis. Curr HIV/AIDS Rep. 2014;11(3):291–307. pmid:25038748
- 34. Wasti SP, van Teijlingen E, Simkhada P, Randall J, Baxter S, Kirkpatrick P, et al. Factors influencing adherence to antiretroviral treatment in Asian developing countries: a systematic review. Trop Med Int Health. 2012;17(1):71–81. pmid:21967241
- 35. Maeda U, Shen B-J, Schwarz ER, Farrell KA, Mallon S. Self-efficacy mediates the associations of social support and depression with treatment adherence in heart failure patients. Int J Behav Med. 2013;20(1):88–96. pmid:22212607
- 36. Friedman MR, Coulter RWS, Silvestre AJ, Stall R, Teplin L, Shoptaw S, et al. Someone to count on: social support as an effect modifier of viral load suppression in a prospective cohort study. AIDS Care. 2017;29(4):469–80. pmid:27456040
- 37. Holt-Lunstad J, Smith TB, Layton JB. Social relationships and mortality risk: a meta-analytic review. PLoS Med. 2010;7(7):e1000316. pmid:20668659
- 38. Bandura A, Freeman WH, Lightsey R. Self-efficacy: the exercise of control. J Cogn Psychother. 1999;13(2):158–66.
- 39. Mi T, Li X, Zhou G, Qiao S, Shen Z, Zhou Y. HIV Disclosure to Family Members and Medication Adherence: Role of Social Support and Self-efficacy. AIDS Behav. 2020;24(1):45–54. pmid:30863978
- 40. Diiorio C, McCarty F, Depadilla L, Resnicow K, Holstad MM, Yeager K, et al. Adherence to antiretroviral medication regimens: a test of a psychosocial model. AIDS Behav. 2009;13(1):10–22. pmid:17978868
- 41. Tassiopoulos K, Huo Y, Kacanek D, Malee K, Nichols S, Mellins CA, et al. Association of Perceived Social Support with Viral Suppression Among Young Adults with Perinatally-Acquired HIV in the US-based Pediatric HIV/AIDS Cohort Study (PHACS). Clin Epidemiol. 2023;15:601–11. pmid:37193342
- 42. Brown WE, Malagala H, Bajunirwe F. Social support, gender and pill burden influence viral load suppression among HIV-infected adolescents and young adults in rural south-western Uganda. Vulnerable Child Youth Stud. 2021;16(1):86–97. pmid:39544698
- 43. Spence AB, Michel K, Wang C, Dutton MA, Lee K, Merenstein D, et al. Viral Suppression Is Associated with HIV Treatment Self-Efficacy in a Cohort of Women in Washington, DC. AIDS Patient Care STDS. 2021;35(3):75–83. pmid:33689457