Healthcare workers (HCWs) are among high-risk subpopulations with a disproportionate burden of mental disorders compared with the general population. Reference Bouaddi, Abdallahi, Fadel Abdi, Hassouni, Jallal and Benjelloun1,Reference Mohammadi, Neyazi, Rangelova, Padhi, Odey and Ogbodum2 Depression, anxiety and post-traumatic stress disorder (PTSD) are particularly prevalent in this population. Reference Li, Scherer, Felix and Kuper3 For instance, evidence from 65 studies included in a systematic review to assess the prevalence of these disorders among a total of 97 333 HCWs globally showed that the pooled prevalence of moderate depression, anxiety and PTSD was 21.7, 22.1 and 21.5%, respectively. Reference Li, Scherer, Felix and Kuper3
Numerous factors contribute to the elevated burden of mental disorders among HCWs. The nature of their work, which often includes long working hours and working night shifts, with little time for social and personal life, contributes significantly to this burden. Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 HCWs also frequently experience sleep problems such as sleep deprivation, insomnia and poor-quality sleep, which are associated with mental health problems. Reference Herrero San Martin, Parra Serrano, Diaz Cambriles, Arias Arias, Muñoz Méndez and Del Yerro Álvarez5,Reference Shumye, Kendall-Tackett and Kassaw6 HCWs are also constantly exposed to human suffering and death in their line of work. Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Muliira, Sendikadiwa and Lwasampijja8 They may also be frequently exposed to abuse, hostility and violence in their workplace. Reference Lua, de Araújo, Santos and de Almeida9 A higher risk of infection with infectious diseases also exacerbates the risk of mental disorders in this population. Reference Hill, Harris, Danielle, Boland, Doherty Alison and Benedetto10 Occasionally being on the front lines dealing with mass casualty events, such as disasters and pandemics like COVID-19, also plays a significant role in the heightened risk of mental disorders in this population. Reference Naushad, Bierens, Nishan, Firjeeth, Mohammad and Maliyakkal11,Reference Stuijfzand, Deforges, Sandoz, Sajin, Jaques and Elmers12
In sub-Saharan Africa (SSA) the burden of mental disorders among HCWs is significant, and is even thought to be higher when compared with other regions. For instance, a systematic review that compared the pooled prevalence of depression among HCWs across regions reported the highest prevalence in Africa (82.4%), compared with pooled prevalence estimates of 33.4, 31.3 and 19.1% in North America, Europe and Asia, respectively. Reference Rezaei, Hoseinipalangi, Rafiei, Dolati, Hosseinifard and Asl13 The burden of mental disorders is thought to be particularly high among sub-Saharan HCWs because they face additional challenges and risk factors such as huge staff shortages, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 high patient:health worker ratios, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 high workloads, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 under-resourced healthcare systems, Reference Kim, Mazenga, Yu, Simon, Nyasulu and Kazembe14 low remuneration and disproportionately high disease burden, including highly infectious epidemics such as Ebola virus disease and HIV. Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15
Despite the high burden of mental disorders among HCWs in SSA, to our knowledge only two systematic reviews assessing these disorders in this region have been published. Reference Chen, Farah, Dong, Chen, Xu and Yin16,Reference Oyat, Oloya, Atim, Ikoona, Aloyo and Kitara17 However, these reviews focused only on studies that were conducted during the COVID-19 pandemic, and consequently they lack data from the periods preceding and following the pandemic. The current review, therefore, has the additional value of providing a picture of mental disorders in this population preceding and following the COVID-19 pandemic. Additionally, while recognising the role of pandemics and other public health emergencies in heightening of mental health problems among HCWs, Reference Naushad, Bierens, Nishan, Firjeeth, Mohammad and Maliyakkal11,Reference Stuijfzand, Deforges, Sandoz, Sajin, Jaques and Elmers12,Reference Søvold, Naslund, Kousoulis, Saxena, Qoronfleh and Grobler18 we also aim to highlight the potential impact of the COVID-19 pandemic on the mental health of HCWs in this region. While this evidence has been provided by the aforementioned reviews, these were conducted in the early stages of the pandemic and thus provide evidence from the early phase only. This review also includes studies from the later phases of the pandemic and utilises the data from these studies to strengthen the evidence on the impact of the pandemic on the mental health of HCWs.
Other existing relevant reviews have been global in nature, Reference Li, Scherer, Felix and Kuper3,Reference Hill, Harris, Danielle, Boland, Doherty Alison and Benedetto10,Reference Rezaei, Hoseinipalangi, Rafiei, Dolati, Hosseinifard and Asl13,Reference Ghahramani, Kasraei, Hayati, Tabrizi and Marzaleh19–Reference Saragih, Tonapa, Saragih, Advani, Batubara and Suarilah21 with the included studies originating mostly outside SSA, thus limiting their generalisability to this context. This review, therefore, aims to systematically summarise the available evidence on the prevalence and factors associated with mental disorders – specifically depression, anxiety and PTSD, or their symptoms – among HCWs from SSA.
Method
Review protocol and registration
Prior to the commencement of this systematic review, we developed the review protocol and registered it with the International Prospective Register of Systematic Reviews (PROSPERO) under registration no. CRD42022349136.
Data sources
We conducted a literature search in the following electronic databases: African Index Medicus, African Journals Online, CINAHL, PsycINFO and PubMed. The search was for articles conducted from database inception to the final day of the database search. Initially, a search had been conducted and concluded on 17 June 2022. This search was later updated to 15 March 2023. A final update was conducted before submission, with the final database search occurring on 15 February 2024. The search was restricted to those studies published in English. All the identified articles were retrieved and uploaded to Rayyan software Reference Ouzzani, Hammady, Fedorowicz and Elmagarmid22 (release year: 2014, accessed via Windows; Rayyan Systems, Inc., Cambridge, Massachusetts, USA; https://www.rayyan.ai/) for data management. In addition, hand-searching was conducted through the reference lists of the included articles and related systematic reviews for additional relevant articles meeting the inclusion criteria of this review.
Search strategy
We used a pre-developed search strategy to search the electronic databases mentioned above. The search strategy contained keywords including ‘depression/anxiety/PSTD’, ‘healthcare workers’ and ‘SSA’, combined using the Boolean Operator ‘AND’. Respective synonyms for these key words were combined using the Boolean operator ‘OR’. Supplementary File 1 provides the search strategy used in the PubMed database.
Eligibility criteria
Table 1 provides the study eligibility criteria for this review.
Table 1 Study eligibility criteria
Screening of articles
P.W. and S.N. independently screened the articles for eligibility. E.K.T. later screened those articles that were identified in the updated search. Articles were screened on three levels, starting with screening for eligibility by title, then by abstract and finally by full text, during which ineligible studies were systematically excluded. Discrepancies were resolved through discussion and consensus with the other study co-authors.
Data extraction
Data were extracted from the selected studies using a standardised, preprepared data extraction form. Data extraction began on 22 August 2022. Data extraction was dually and independently carried out by P.W. and S.N.; E.K.T. extracted data from the final updated articles. The extracted data include the following: (a) study details (name of first author, year of publication, country of origin and study design); (b) study participant characteristics (cadre of healthcare personnel recruited, sample size, sampling methods, age and gender); and (c) study outcomes (prevalence of depression, anxiety or PTSD, measurement tools used, cut-off scores applied (for screening tools), psychometric information of the tools used and factors associated with depression, anxiet, and PTSD among HCWs (alongside the reported measure of effect and precision estimate). Where the data of interest were not available in an article, the corresponding author was contacted to clarify or provide missing data. Papers where the contacted authors did not respond to the request were excluded from the review.
Quality appraisal
The quality of the included articles was assessed using the Newcastle–Ottawa Scale (NOS) quality assessment scale adapted for cross-sectional studies. P.W. and S.N. equally shared the included studies and independently conducted quality appraisal. To reduce bias, E.K.T. conducted an independent appraisal of all the articles and compared scores with the individual scores from P.W. and S.N. Disagreements in quality ratings were resolved through discussion and consensus. The NOS uses a star scoring system with a maximum of nine points assigned to each article. This quality assessment tool assesses three main domains, including the ascertainment outcome of interest, the selection of study participants and the comparability of study groups. The scoring system awards five points for the selection domain, one point for the comparability domain and three points for the outcome domain. Scores of 0–4, 5–6, 7–8 and 9 are indicative of unsatisfactory, satisfactory, good and very good quality, respectively. In this review we used a score of ≥7 to indicate high-quality studies.
Data synthesis
Because the significant heterogeneity of the measurement tools used across the included studies precluded a meta-analysis, we therefore conducted a narrative synthesis. We narratively summarised the prevalence estimates of depression, anxiety and PTSD among HCWs in SSA, as well as their associated factors, by the investigated outcome of interest. In this review, only those factors significantly associated with the outcomes of interest (depression, anxiety or PTSD) at P < 0.05 in the multivariable analysis were considered and extracted.
Results
Results of database search
The database search yielded a total of 4832 articles (African Index Medicus, n = 44; African Journal Online, n = 369; CINAHL, n = 428; PsycINFO, n = 2141; and PubMed, n = 1850). Ten additional articles were retrieved from the reference lists of the included articles and related systematic reviews. Following the removal of duplicates and screening of articles based on the eligibility criteria, 69 articles were included in this review. Figure 1 shows the PRISMA flowchart for the systematic review process.
Fig. 1 PRISMA flowchart for the systematic review process.
Characteristics of included studies
Supplementary File 2 presents in detail the characteristics of the 69 studies included in this review; in summary, all included studies were cross-sectional in design. The included studies were conducted across 27 countries in SSA, with more than half (n = 37, 53.6%) being conducted in only three countries in SSA (Ethiopia, Nigeria and South Africa) (Fig. 2). Four of the included studies Reference Assefa, Soura, Hemler, Korte, Wang and Abdullahi23–Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26 were multi-country studies that recruited HCWs from either 16, Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24 13, Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26 3 Reference Assefa, Soura, Hemler, Korte, Wang and Abdullahi23 or 2 countries. Reference Olashore, Akanni and Oderinde25 The total number of HCWs across these studies was 24 266. A comparison sample of 502 participants from the general population was also recruited in one of the included studies. Reference Agberotimi, Akinsola, Oguntayo and Olaseni27
Fig. 2 Distribution of the included studies.
Almost 80% of the included studies (n = 55, 79.7%) were studies assessing the burden of depression, anxiety or PTSD among HCWs in the context of the COVID-19 pandemic. The remaining 14 studies Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Muliira, Sendikadiwa and Lwasampijja8,Reference Kim, Mazenga, Yu, Simon, Nyasulu and Kazembe14,Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24,Reference Aliyu and Adeniyi28–Reference Pindar, Wakil, Coker and Abdul36 were conducted prior to the onset of the pandemic. Most of the included studies (n = 48, 69.6%) were conducted in health facilities while the remainder (n = 21, 30.4%) were conducted virtually through online platforms.
Over 70% of the included studies (n = 50, 72.5%) recruited multiple cadres of HCWs as study participants. In the remaining 19 studies, specific cadres of HCWs were recruited, including nurses recruited in 8 studies, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Olabisi, Dosumu, Oyewumi, Adegoke, Oladotun and Aremu34,Reference Ali, Shah and Talib37–Reference Vancampfort and Mugisha41 doctors and nurses recruited in 3 studies, Reference Ibigbami, Akinsulore, Opakunle, Seun-Fadipe, Oginni and Okorie42–Reference Workneh, Worku, Assefa and Berhane44 doctors in 5 studies, Reference Bernard Ubom, Adebayo, Adeoye, Kanmodi, Salihu and Umar30,Reference Naidoo, Tomita and Paruk32,Reference Ali, Shah, Du, Leekha and Talib45–Reference Hain, Tomita, Milligan and Chiliza47 surgeons in 1 study, Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24 health extension workers in 1 study Reference Birhane, Medhin, Demissie, Tassew, Gebru and Tadesse31 and midwives in 1 study. Reference Muliira, Sendikadiwa and Lwasampijja8 Study participants were recruited non-randomly in nearly half of the included studies (n = 28, 40.6%), whereas 42.0% (n = 29) of the included studies recruited their study participants using random sampling techniques. In the remaining 12 studies, information on the sampling methods was not reported.
Most of the included studies (n = 40, 58.0%) concurrently assessed at least two or more of the mental disorders of interest in this review (depression, anxiety and PTSD). In the remaining 29 studies, only 1 mental disorder was assessed, including depression in 14 studies, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Kim, Mazenga, Yu, Simon, Nyasulu and Kazembe14,Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24,Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26,Reference Aliyu and Adeniyi28,Reference Belete and Anbesaw29,Reference Birhane, Medhin, Demissie, Tassew, Gebru and Tadesse31,Reference Obi, Aniebue, Okonkwo, Okeke and Ugwunna33,Reference Pindar, Wakil, Coker and Abdul36,Reference Workneh, Worku, Assefa and Berhane44,Reference Wayessa, Melesse, Amaje Hadona and Wako48–Reference Simbeza, Mutale, Mulabe, Jere, Bukankala and Sikombe51 anxiety in 9 studies Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Muliira, Sendikadiwa and Lwasampijja8,Reference Olashore, Akanni and Oderinde25,Reference Bernard Ubom, Adebayo, Adeoye, Kanmodi, Salihu and Umar30,Reference Chorwe-Sungani38,Reference Kibret, Teshome, Fenta, Hunie and Tamire52–Reference Wayessa, Melesse and Hadona55 and PTSD in 7 studies. Reference Olashore, Akanni, Molebatsi and Ogunjumo35,Reference Kabunga and Okalo39,Reference Vancampfort and Mugisha41,Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56–Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59 None of the included studies used a diagnostic interview to diagnose any of these disorders. Rather, all the included studies used mental health screening tools to identify the symptoms of each disorder. Information on the reliability and/or validity of these screening tools for use among HCWs from SSA is limited. Many of the included studies (n = 40, 57.1%) did not report any of this information. Where reported, only reliability (Cronbach’s α) was reported across studies (Table 2).
Table 2 Prevalence estimates of depression, anxiety and post-traumatic stress disorder (PTSD) among healthcare workers (HCWs) from sub-Saharan Africa (SSA)
PC-PTSD (DSM-V), Primary Care – Posttraumatic Stress Disorder for Diagnostic Statistical Manual (DSM) V; PCL-C-PTSD, Checklist – Civilian Version; PCL-5-PTSD, Checklist for DSM-5; PSS, Perceived Stress Scale; STAI, State Trait Anxiety Scale; DTS, Davidson Trauma Scale; SARSQ, Stanford Acute Reaction Stress Questionnaire; PSS-SR, Symptom Scale for Posttraumatic Stress Disorder – Self Reporting Version; HARS, Hamilton Anxiety Rating Scale; DASS, Depression, Anxiety and Stress Scale; IES-R, Revised Impact of Event Scale; HAD, Hospital Depression and Anxiety Scale; PHQ, Patient Health Questionnaire; GAD, generalised anxiety disorder; BDI-II, Beck’s Depression Inventory-II; CAS, COVID-19 Anxiety Scale; SRQ, Self-Reporting Questionnaire; HSCL, Hopkin’s Symptoms Checklist; NR, not reported; NA, not assessed.
Prevalence of depression, anxiety and PTSD among HCWs from SSA
All the included studies reported the prevalence of one or more of the mental disorders of interest in this review (depression, anxiety and PTSD). Across these studies, wide-ranging prevalence estimates of depression, anxiety or PTSD among HCWs from SSA were reported. Table 2 summarises the prevalence of each of these disorders as reported in these studies, including the measures used and the cut-off scores applied, where reported.
Prevalence of depressive symptoms among HCWs from SSA
Fifty-two studies reported the prevalence of depressive symptoms, either exclusively or concurrently with other mental disorders. Across these studies, depressive symptoms were assessed using several measurement tools, including various versions of the Patient Health Questionnaire Reference Kroenke, Spitzer and Williams86 used in 29 studies, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Assefa, Soura, Hemler, Korte, Wang and Abdullahi23,Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24,Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26,Reference Agberotimi, Akinsola, Oguntayo and Olaseni27,Reference Belete and Anbesaw29,Reference Birhane, Medhin, Demissie, Tassew, Gebru and Tadesse31,Reference Naidoo, Tomita and Paruk32,Reference Ali, Shah and Talib37,Reference Ibigbami, Akinsulore, Opakunle, Seun-Fadipe, Oginni and Okorie42–Reference Ali, Shah, Du, Leekha and Talib45,Reference Hain, Tomita, Milligan and Chiliza47,Reference Yadeta, Dessie and Balis49,Reference Simbeza, Mutale, Mulabe, Jere, Bukankala and Sikombe51,Reference Bundi, Poipoi and Morema65,Reference Burnett-Zieman, Warren, Chiundira, Mandala, Kachale and McHoma66,Reference Elamin, Hamza, Abdalla, Mustafa, Altayeb and Mohammed68,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Hassan Salman, Zainelabdin Mohamed Elmahdi and Elnour72,Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75–Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78,Reference Oguntayo, Akinsola, Olaseni and Agberotimi81–Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 the Depression, Anxiety and Stress Scale Reference Lovibond and Lovibond87 used in 12 studies, Reference Aliyu and Adeniyi28,Reference Olabisi, Dosumu, Oyewumi, Adegoke, Oladotun and Aremu34,Reference Mekonen, Shetie and Muluneh40,Reference Duffton, Heystek, Engelbrecht, Rajan and Du Toit46,Reference Wayessa, Melesse, Amaje Hadona and Wako48,Reference Ahmed, Miskeen, Awadelgeed and Al Faifi60,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Dawood, Tomita and Ramlall67,Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74,Reference Ofori, Osarfo, Agbeno, Manu and Amoah80,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 Beck’s Depression Inventory Reference Beck, Ward, Mendelson, Mock and Erbaugh88 used in 3 studies, Reference Pindar, Wakil, Coker and Abdul36,Reference Arthur-Mensah, Paintsil, Agudu Delali and Kyei62,Reference Bapolisi, Maurage, Rubambura, Tumaini, Baguma and Cikomola64 the Hospital Anxiety and Depression Scale Reference Zigmond and Snaith89 used in 5 studies, Reference Shumye, Kendall-Tackett and Kassaw6,Reference Ariyo, Akinnawo, Akpunne, Kumuyi and Onisile61,Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69,Reference Idrees and Bashir73,Reference Nguépy Keubo, Mboua, Djifack Tadongfack, Fokouong Tchoffo, Tasson Tatang and Ide Zeuna79 the Self-Reporting Questionnaire 90 used in 2 studies Reference Kim, Mazenga, Yu, Simon, Nyasulu and Kazembe14,Reference Phiri, Songo, Whitehead, Chikuse, Moucheraud and Dovel50 and the Zung Self-Rating Depression Scale Reference Zung91 used in 1 study. Reference Obi, Aniebue, Okonkwo, Okeke and Ugwunna33
Wide-ranging prevalence estimates of depressive symptoms were reported across studies (Table 2). Regardless of the measurement tool used, the prevalence of depression ranged from 2.1 to 75.7%. One study from Nigeria Reference Agberotimi, Akinsola, Oguntayo and Olaseni27 compared the prevalence of depressive symptoms between HCWs and the general population and reported a significantly higher prevalence among HCWs (35.1 versus 23.5%, P < 0.01). One study that was conducted across 13 countries compared the prevalence of daily depressive symptoms before and during the COVID-19 pandemic. Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26 In that study, the prevalence of daily depressive symptoms among 439 HCWs before and during the pandemic was 2.1 and 20.0%, respectively.
Prevalence of anxiety symptoms among HCWs from SSA
Forty-five studies assessed the prevalence of anxiety symptoms, either exclusively or concurrently with other mental disorders. Across these studies, anxiety symptoms were assessed using several anxiety symptom screeners, including, the 7-item Generalised Anxiety Disorder Scale Reference Spitzer, Kroenke, Williams and Löwe92 used in 22 studies, Reference Agberotimi, Akinsola, Oguntayo and Olaseni27,Reference Bernard Ubom, Adebayo, Adeoye, Kanmodi, Salihu and Umar30,Reference Naidoo, Tomita and Paruk32,Reference Ali, Shah and Talib37,Reference Ibigbami, Akinsulore, Opakunle, Seun-Fadipe, Oginni and Okorie42,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Ali, Shah, Du, Leekha and Talib45,Reference Hain, Tomita, Milligan and Chiliza47,Reference Kibret, Teshome, Fenta, Hunie and Tamire52–Reference Osasona and Oderinde54,Reference Bundi, Poipoi and Morema65,Reference Elamin, Hamza, Abdalla, Mustafa, Altayeb and Mohammed68,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Hassan Salman, Zainelabdin Mohamed Elmahdi and Elnour72,Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75–Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78,Reference Oguntayo, Akinsola, Olaseni and Agberotimi81,Reference Onchonga, Ngetich, Makunda, Wainaina, Wangeshi and Viktoria83,Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 the Depression, Anxiety and Stress Scale Reference Lovibond and Lovibond87 used in 11 studies, Reference Olabisi, Dosumu, Oyewumi, Adegoke, Oladotun and Aremu34,Reference Mekonen, Shetie and Muluneh40,Reference Duffton, Heystek, Engelbrecht, Rajan and Du Toit46,Reference Wayessa, Melesse and Hadona55,Reference Ahmed, Miskeen, Awadelgeed and Al Faifi60,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Dawood, Tomita and Ramlall67,Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74,Reference Ofori, Osarfo, Agbeno, Manu and Amoah80,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 the Anxiety Rating Scale Reference Komor93 used in 2 studies, Reference Olashore, Akanni and Oderinde25,Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82 the Hospital Anxiety and Depression Scale Reference Zigmond and Snaith89 used in 4 studies, Reference Ariyo, Akinnawo, Akpunne, Kumuyi and Onisile61,Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69,Reference Idrees and Bashir73,Reference Nguépy Keubo, Mboua, Djifack Tadongfack, Fokouong Tchoffo, Tasson Tatang and Ide Zeuna79 a 4-item version of the Patient Health Questionnaire Reference Kroenke, Spitzer and Williams86 used in 1 study, Reference Assefa, Soura, Hemler, Korte, Wang and Abdullahi23 the State Trait Anxiety Inventory Reference Spielberger94 used in 1 study, Reference Arthur-Mensah, Paintsil, Agudu Delali and Kyei62 the Coronavirus Anxiety Scale Reference Lee95 used in 2 studies, Reference Chorwe-Sungani38,Reference Mc Magh, Fadahun and Francis76 the Hopkins Symptom Checklist Reference Derogatis, Lipman, Rickels, Uhlenhuth and Covi96 used in 1 study, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 the Hamilton Anxiety Rating Scale Reference Hamilton97 used in 1 study Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7 and the Death Anxiety Subscale of the Death Distress Scale Reference Abdel-Khalek98 used in 1 study. Reference Naidoo, Tomita and Paruk32
Across the 45 studies, wide-ranging prevalence estimates of anxiety symptoms, including symptoms of generalised anxiety disorder, trait anxiety, state anxiety, COVID-19 anxiety and death anxiety, were reported (Table 2). Overall, the prevalence of anxiety symptoms ranged from 4.8 to 96.5%. One study compared the symptoms of generalised anxiety disorder among 382 HCWs and 502 participants from the general population, and reported a prevalence of 58.4% and 49.6% among HCWs and the general population, respectively. Reference Agberotimi, Akinsola, Oguntayo and Olaseni27
Prevalence of PTSD symptoms among HCWs from SSA
In total, 21 of the included studies reported the prevalence of PTSD symptoms among HCWs from SSA. Across these studies, PTSD symptoms were assessed using various PTSD symptom screeners, including the revised Impact of Event Scale Reference Weiss99 used in 12 studies, Reference Agberotimi, Akinsola, Oguntayo and Olaseni27,Reference Ali, Shah and Talib37,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Ali, Shah, Du, Leekha and Talib45,Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57,Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59,Reference Dawood, Tomita and Ramlall67,Reference Elamin, Hamza, Abdalla, Mustafa, Altayeb and Mohammed68,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71,Reference Oguntayo, Akinsola, Olaseni and Agberotimi81 the PTSD Checklist for Diagnostic Statistical Manual 5 (DSM-5) Reference Blevins, Weathers, Davis, Witte and Domino100 used in 3 studies, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15,Reference Vancampfort and Mugisha41,Reference Human, Vahed and Marais58 the PTSD Checklist – Civilian Version Reference Weathers, Huska and Keane101 used in 2 studies, Reference Olashore, Akanni, Molebatsi and Ogunjumo35,Reference Kabunga and Okalo39 the Primary Care – Posttraumatic Stress Disorder for DSM-5 Reference Prins, Bovin, Smolenski, Marx, Kimerling and Jenkins-Guarnieri102 used in 2 studies, Reference Burnett-Zieman, Warren, Chiundira, Mandala, Kachale and McHoma66,Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 the Stanford Acute Reaction Stress Questionnaire Reference Cardeña, Koopman, Classen, Waelde and Spiegel103 used in 1 study Reference Bapolisi, Maurage, Rubambura, Tumaini, Baguma and Cikomola64 and the Symptom Scale for Posttraumatic Stress Disorder – Self-Reporting Version Reference Foa, Riggs, Dancu and Rothbaum104 used in 1 study. Reference Shumye, Kendall-Tackett and Kassaw6
Similar to depressive and anxiety symptoms, wide-ranging prevalence estimates of PTSD symptoms were also reported across studies (Table 2). The prevalence of PTSD symptoms among HCWs as reported across these studies ranged from 11.7 to 78.3%. One study from Nigeria compared the prevalence of PTSD symptoms among 382 HCWs and 502 participants from the general population, and reported a significantly higher prevalence among HCWs (52.6 versus 42.8%, P < 0.05). Reference Agberotimi, Akinsola, Oguntayo and Olaseni27
Correlates of depression, anxiety and PTSD among HCWs from SSA
In total, 36 of the included studies reported the correlates of one or more of the mental disorders of interest in this review (depression, anxiety and PTSD). Tables 3, 4 and 5 present detailed summaries of these correlates, including their effect sizes. Across these studies, several correlates of depression, anxiety or PTSD among HCWs from SSA were reported. Overall, these correlates can be classified into sociodemographic, health-related, COVID-19-related and work-related correlates, and are reported as such in this paper.
Table 3 Correlates of depression among healthcare workers from sub-Saharan Africa
β, beta coefficient (adjusted); AOR, adjusted odds ratio; IRR, incidence rate ratio; NR, none reported.
Table 4 Correlates of anxiety among healthcare workers from sub-Saharan Africa
β, beta coefficient (adjusted); AOR, adjusted odds ratio; APOR, adjusted proportional odds ratio; IRR, incidence rate ratio; PTSD, post-traumatic stress disorder; EVD, Ebola virus disease; PPE, personal protective equipment; NR, none reported.
Table 5 Correlates of post-traumatic stress disorder among healthcare workers from sub-Saharan Africa
β, beta coefficient (adjusted); AOR, adjusted odds ratio; EVD, Ebola virus disease; PPE, personal protective equipment; NR, none reported.
Correlates of depressive symptoms among HCWs from SSA
Twenty-one of the included studies Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Belete and Anbesaw29,Reference Birhane, Medhin, Demissie, Tassew, Gebru and Tadesse31,Reference Naidoo, Tomita and Paruk32,Reference Mekonen, Shetie and Muluneh40,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43–Reference Ali, Shah, Du, Leekha and Talib45,Reference Wayessa, Melesse, Amaje Hadona and Wako48–Reference Phiri, Songo, Whitehead, Chikuse, Moucheraud and Dovel50,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74–Reference Mc Magh, Fadahun and Francis76,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78,Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82,Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 reported the correlates of depressive symptoms among HCWs (Table 3). The sociodemographic factors that were reported to significantly increase the risk of depressive symptoms among HCWs include older age, Reference Naidoo, Tomita and Paruk32 younger age, Reference Wayessa, Melesse, Amaje Hadona and Wako48,Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 female gender, Reference Belete and Anbesaw29,Reference Naidoo, Tomita and Paruk32,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Workneh, Worku, Assefa and Berhane44,Reference Yadeta, Dessie and Balis49,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74–Reference Mc Magh, Fadahun and Francis76,Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 being unmarried, Reference Belete and Anbesaw29,Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 being married, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63 living alone, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63 male gender, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 higher education, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 lower education, Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 living with family, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 having a higher number of economically dependent family members, Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 being a Christian Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 and substance use. Reference Belete and Anbesaw29,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 Older age was the only sociodemographic factor reported to be protective against depressive symptoms, reported in only one study. Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82
Among health-related factors reported to increase the risk of depressive symptoms were having a chronic medical illness, Reference Mekonen, Shetie and Muluneh40,Reference Wayessa, Melesse, Amaje Hadona and Wako48 history of mental illness, Reference Belete and Anbesaw29,Reference Mekonen, Shetie and Muluneh40 neuroticism Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82 and burnout. Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4,Reference Birhane, Medhin, Demissie, Tassew, Gebru and Tadesse31,Reference Naidoo, Tomita and Paruk32,Reference Phiri, Songo, Whitehead, Chikuse, Moucheraud and Dovel50 Resilience was the only health-related factor reported to lower the risk of depressive symptoms among HCWs, reported in one study. Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82
Several COVID-19-related factors were reported to increase the risk of depressive symptoms among HCWs, including working on the front lines, Reference Ali, Shah and Talib37,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 being infected with COVID-19, Reference Workneh, Worku, Assefa and Berhane44,Reference Phiri, Songo, Whitehead, Chikuse, Moucheraud and Dovel50 expecting to be infected with COVID-19 within the next 12 months, Reference Phiri, Songo, Whitehead, Chikuse, Moucheraud and Dovel50 higher perceived susceptibility to COVID-19, Reference Yadeta, Dessie and Balis49 lack of knowledge on COVID-19, Reference Wayessa, Melesse, Amaje Hadona and Wako48 not having guidelines on COVID-19 management, Reference Mekonen, Shetie and Muluneh40,Reference Workneh, Worku, Assefa and Berhane44 negative feedback from family and friends regarding being on the front lineS, Reference Mekonen, Shetie and Muluneh40 not satisfied with government support during the pandemic, Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 working in a COVID-19 treatment centre Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 and experiencing stigma due to working on the front lines. Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 Having training in COVID-19 management, Reference Wayessa, Melesse, Amaje Hadona and Wako48 having personal protective equipment Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 and working in the largest COVID-19 isolation centre Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 were the only COVID-19-related factors reported to be protective against depressive symptoms among HCWs during the pandemic.
Among work-related factors reported to increase the risk of depressive symptoms were having greater numbers of night shifts per week, Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 being a nurse, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63 being a pharmacist, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 being a physician, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 being a doctor, Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43 being a medical laboratory technologist, Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 having a greater perception of moral transgression in the workplace, Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 working in centres offering consultations to the general population as opposed to only people living with HIV Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 and working outside the capital city. Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 Only four work-related factors were reported to be protective against depressive symptoms, namely having more years of experience, Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 being a nurse as opposed to other cadres of HCWs, Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 a higher nurse:patient ratio Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 and working in an in-patient department. Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63
Correlates of anxiety symptoms among HCWs from SSA
In total, 21 of the included studies Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Muliira, Sendikadiwa and Lwasampijja8,Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15,Reference Olashore, Akanni and Oderinde25,Reference Naidoo, Tomita and Paruk32,Reference Mekonen, Shetie and Muluneh40,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Kibret, Teshome, Fenta, Hunie and Tamire52,Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53,Reference Wayessa, Melesse and Hadona55,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Dawood, Tomita and Ramlall67,Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74–Reference Mc Magh, Fadahun and Francis76,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78,Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82,Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 reported the correlates of anxiety symptoms among HCWs from SSA (Table 4). Across these studies, the sociodemographic factors that were reported to increase the risk of anxiety symptoms include older age, Reference Naidoo, Tomita and Paruk32,Reference Kibret, Teshome, Fenta, Hunie and Tamire52,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 younger age, Reference Wayessa, Melesse and Hadona55 being married, Reference Kibret, Teshome, Fenta, Hunie and Tamire52,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63 female gender, Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63,Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74–Reference Mc Magh, Fadahun and Francis76 being unmarried, Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 male gender, Reference Dawood, Tomita and Ramlall67 higher education, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 lower income, Reference Wayessa, Melesse and Hadona55,Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 having household members with chronic heart or lung disease, Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 being of non-African descent Reference Mc Magh, Fadahun and Francis76 and alcohol or other drug use. Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Wayessa, Melesse and Hadona55 On the other hand, low education Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82 and social support Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15,Reference Olashore, Akanni and Oderinde25 were the only reported protective factors of anxiety symptoms.
The health-related factors reported to be associated with a higher risk of anxiety symptoms were having a chronic illness, Reference Mekonen, Shetie and Muluneh40,Reference Kibret, Teshome, Fenta, Hunie and Tamire52,Reference Wayessa, Melesse and Hadona55 neuroticism, Reference Olashore, Akanni and Oderinde25,Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82 burnout, Reference Naidoo, Tomita and Paruk32 having a history of mental illness, Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 exposure to Ebola virus disease, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 experiencing Ebola virus disease stigma, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 coping with maternal death using planning, active coping or acceptance as coping strategies, Reference Muliira, Sendikadiwa and Lwasampijja8 lack of confidence in coping with stress Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 and having a lower perceived health status. Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 Resilience, Reference Olashore, Akanni and Oderinde25,Reference Olashore, Molebatsi, Musindo, Bojosi, Obadia and Molefe-Baikai82 not feeling overwhelmed by the demands of everyday life Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 and not feeling that you cannot make it Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 were the health-related factors reported to lower the risk of anxiety symptoms.
The COVID-19-related factors reported to increase the risk of anxiety symptoms include having a family member with suspected COVID-19, Reference Kibret, Teshome, Fenta, Hunie and Tamire52 lack of access to personal protective equipment, Reference Kibret, Teshome, Fenta, Hunie and Tamire52 lack of COVID-19 management guidelines, Reference Mekonen, Shetie and Muluneh40 having infected family members, Reference GebreEyesus, Tarekegn, Amlak, Shiferaw, Emeria and Geleta70 working on the front lines, Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 working in a COVID-19 treatment centre, Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 losing relatives or friends to COVID-19, Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 contact with suspected or confirmed case of COVID-19, Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 lack of COVID-19 updates Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 and COVID-19-related worry. Reference Teshome, Glagn, Shegaze, Tekabe, Getie and Assefa53 Working in the largest COVID-19 isolation centre Reference Siamisang, Kebadiretse, Tjirare, Muyela, Gare and Masupe85 and having personal protective equipment Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 were the only COVID-19-related factors reported to be protective against anxiety symptoms during the pandemic.
Among work-related factors reported to increase the risk of anxiety symptoms among HCWs were working in a private facility, Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 being a nurse, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar63 being a doctor, Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43 being a health worker, Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 working outside the capital city, Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 being a medical laboratory technologist, Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 work overload, Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7 working night shifts, Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7 witnessing two or more maternal deaths in the past 2 years, Reference Muliira, Sendikadiwa and Lwasampijja8 being in charge of four or more maternal deaths, Reference Muliira, Sendikadiwa and Lwasampijja8 having a greater perception of moral transgression in the workplace Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 and lack of training in handling death situations. Reference Muliira, Sendikadiwa and Lwasampijja8 Having more years of experience Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 and a higher nurse:patient ratio Reference Sagaon-Teyssier, Kamissoko, Yattassaye, Diallo, Rojas Castro and Delabre84 were the only reported protective factors against anxiety symptoms.
Correlates of PTSD symptoms among HCWs from SSA
Eleven of the included studies Reference Shumye, Kendall-Tackett and Kassaw6,Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15,Reference Olashore, Akanni, Molebatsi and Ogunjumo35,Reference Kabunga and Okalo39,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57,Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59,Reference Dawood, Tomita and Ramlall67,Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 reported the correlates of PTSD symptoms among HCWs from SSA (Table 5). Across these studies, the reported sociodemographic risk factors of PTSD symptoms include female gender, Reference Shumye, Kendall-Tackett and Kassaw6,Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 being married, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 male gender, Reference Dawood, Tomita and Ramlall67 substance use, Reference Shumye, Kendall-Tackett and Kassaw6,Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71 older age, Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 younger age Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59 and poor social support. Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 Higher social support was the only reported sociodemographic protective factor of PTSD symptoms. Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15,Reference Kabunga and Okalo39
The health-related factors that were reported to increase the risk of PTSD symptoms include having a chronic illness, Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 neuroticism, Reference Olashore, Akanni, Molebatsi and Ogunjumo35 insomnia, Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59 depression, Reference Hajure, Dibaba, Shemsu, Desalegn, Reshad and Mohammedhussein71 exposure to Ebola virus disease, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 experience of Ebola virus disease stigma, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 poor sleep quality, Reference Shumye, Kendall-Tackett and Kassaw6 experiencing stigma associated with being a health worker Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56,Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59 and history of mental illness. Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 No study reported any protective health-related factors of PTSD symptoms.
The reported COVID-19-related risk factors of PTSD symptoms include fear of getting infected, Reference Kabunga and Okalo39 perceiving COVID-19 as a risk, Reference Dawood, Tomita and Ramlall67 working on the front lines, Reference Shah, Monroe-Wise, Talib, Nabiswa, Said and Abeid43,Reference Mulatu, Tesfaye, Woldeyes, Bayisa, Fisseha and Kassu78 COVID-19-related stigma, Reference Cénat, Rousseau, Bukaka, Dalexis and Guerrier15 working in high-risk COVID-19 wards, Reference Shumye, Kendall-Tackett and Kassaw6 not having daily updates on COVID-19 Reference Yitayih, Mekonen, Zeynudin, Mengistie and Ambelu59 and lack of personal protective equipment. Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 Pandemic preparedness was the only reported protective factor against PTSD symptoms among HCWs. Reference Dawood, Tomita and Ramlall67
The only work-related factors reported to increase the risk of PTSD symptoms include being a nurse, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 experiencing workplace violence Reference Olashore, Akanni, Molebatsi and Ogunjumo35 and increased workload. Reference Kabunga and Okalo39 Working with in-patients, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 working in units other than the emergency department, Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 working in the out-patient department Reference Ayalew, Deribe, Abraham, Reta, Tadesse and Defar57 and being a physician Reference Asnakew, Legas, Muche Liyeh, Belete, Haile and Yitbarek56 were the only reported protective factors against PTSD symptoms.
Impact of the COVID-19 pandemic on depression, anxiety and PTSD among HCWs from SSA
From this review, it appears that the COVID-19 pandemic increased the burden and/or risk of depression, anxiety and PTSD among HCWs in SSA. As earlier reported, 55 of the included studies were conducted during the pandemic while the remaining 14 Reference Belayneh, Zegeye, Tadesse, Asrat, Ayano and Mekuriaw7,Reference Muliira, Sendikadiwa and Lwasampijja8,Reference Kim, Mazenga, Yu, Simon, Nyasulu and Kazembe14,Reference Commander, Ellis, Williamson, Grabski, Sallah and Derbew24,Reference Aliyu and Adeniyi28–Reference Pindar, Wakil, Coker and Abdul36 were conducted prior to the onset of the pandemic. Comparing the results from these studies, those conducted during the pandemic appear to show relatively higher prevalence estimates of depression, anxiety and PTSD compared with those conducted before the pandemic. Wide-ranging prevalence estimates were reported for these mental disorders both during and before the pandemic. The prevalence of depression ranged from 2.1 to 62.2% pre-pandemic and 6.5 to 75.7% during the pandemic. Similarly, the prevalence of anxiety ranged from 19.8 to 46.6% pre-pandemic while it ranged from 4.8 to 96.5% in the later phases. The prevalence of PTSD pre-pandemic was 18.4% (only 1 study assessed PTSD pre-pandemic). The prevalence of PTSD ranged from 11.7 to 78.3% during the pandemic.
Despite the wide ranges observed, the reported prevalence of depressive symptoms was as high as 75.7% during the pandemic Reference Jemal, Deriba, Geleta, Tesema, Awol and Mengistu74 compared with 62.2% before. Reference Mbanga, Makebe, Tim, Fonkou, Toukam and Njim4 When comparing the prevalence of depressive symptoms among the same set of HCWs before and during the pandemic, 1 study Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26 reported a higher prevalence during the pandemic (20.0% during the pandemic versus 2.1% pre-pandemic). The prevalence of anxiety symptoms was as high as 96.5% during the pandemic Reference Falade, Oyebanji, Oshatimi, Babatola, Orekoya and Eegunranti69 compared with 46.6% before. Reference Muliira, Sendikadiwa and Lwasampijja8 Finally, the prevalence of PTSD symptoms was as high as 78.3% during the pandemic Reference Kwobah, Mwangi, Patel, Mwogi, Kiptoo and Atwoli75 compared with 18.4% before. Reference Olashore, Akanni, Molebatsi and Ogunjumo35 Regarding COVID-19 increasing the risk of these disorders, many COVID-19-related factors were reported to increase the risk of depression, anxiety and PTSD among HCWs from SSA, with some of these increasing the risk of these disorders by up to nine times (please see the subtopics on the correlates of each disorder and Tables 3, 4 and 5).
Quality of the included studies
The quality scores of the included studies are presented in Supplementary File 3. Overall, 32 of the included studies were rated to be of good quality, 31 of satisfactory quality and 6 of unsatisfactory quality. Based on a cut-off score of ≥7 on the NOS, only 32 of these studies were of high quality while the remaining 37 were of low quality and with a high risk of bias. When comparing the results of high- and low-quality studies, the prevalence of depression as reported by the former ranged from 6.7 to 75.7%, compared with a range of 2.1 to 66.5% by the latter. The prevalence of anxiety as reported by the high-quality studies ranged from 3.2 to 79.0%, compared with 4.8 to 96.5% by the low-quality studies. The prevalence of PTSD as reported by the high-quality studies ranged from 12.0 to 61.7%, compared with 11.7 to 40.2% reported by the low-quality studies.
Discussion
We conducted this systematic review to summarise the available evidence on the burden and factors associated with depression, anxiety and PTSD among HCWs from SSA. A total of 69 studies met our inclusion criteria and were reviewed. Almost 80% of the included studies (n = 55) specifically investigated the burden of depression, anxiety and/or PTSD, and their associated factors, among HCWs in the context of the COVID-19 pandemic. Given the virus’s high rates of transmission and associated mortality, this might highlight the rapid and important response by researchers to understanding the mental health of a key group of the population that was at the forefront of the fight against the pandemic.
In this review, we note that most of the included studies focused on either depression (n = 52) or anxiety (n = 45), with a limited focus on PTSD (n = 21). In part, this may be attributed to the paucity of tools adequately validated for PTSD in this region. Evidence from a systematic review that assessed the availability of validated screening tools for common mental disorders in low- and middle-income countries found that, of the 275 identified validations, only 13 were for PTSD measurement tools compared with 175 for measures of depression and 24 for measures of anxiety. Reference Ali, Ryan and De Silva105 Further studies aimed at developing and/or validating the measures of PTSD among HCWs in this setting are needed.
Wide-ranging prevalence estimates of depression, anxiety and PTSD among HCWs from SSA were reported across the included studies in this review; the prevalence of depressive, anxiety and PTSD symptoms ranged from 2.1 to 75.7, 4.8 to 96.5 and 11.7 to 78.3%, respectively. Despite the wide ranges and the high estimates observed, these prevalence estimates should be interpreted with caution because of several factors that may have resulted in biased estimates. Many of the participants in the included studies were recruited non-randomly, information on the reliability and validity of the measures used in these studies was sparse and most of the included studies were of poor quality. With that said, wide-ranging prevalence estimates of these disorders were also reported in a recent narrative review involving HCWs from SSA during the COVID-19 pandemic. Reference Oyat, Oloya, Atim, Ikoona, Aloyo and Kitara17 Similar wide-ranging findings have also been reported in reviews involving HCWs from across the globe, although these reviews also focused on the COVID-19 pandemic. Reference De Kock, Latham, Leslie, Grindle, Munoz and Ellis106,Reference Muller, Hafstad, Himmels, Smedslund, Flottorp and Stensland107 Differences in sample sizes, study populations, measurement tools, cut-off scores for similar measures and study settings (for example, exposure to different risk factors or differences in exposure levels for similar risk factors) may explain the wide variation in the reported prevalence estimates of depression, anxiety and PTSD among HCWs from SSA. Despite the wide-ranging estimates, the prevalence of these disorders among HCWs from this region appears to be high, even when compared with that in the general population as reported in one of the included studies. Reference Agberotimi, Akinsola, Oguntayo and Olaseni27 Studies from other regions that recruited a comparison group from the general population also report higher estimates among HCWs. Reference Hassannia, Taghizadeh, Moosazadeh, Zarghami, Taghizadeh and Dooki108 However, given that only one study in this review included a comparison group, future studies on the mental health of HCWs should aim to include comparison groups to enable more definite conclusions. That said, the high burden noted here calls for targeted psychosocial interventions to support this key and important group. This is especially important because the existence of these disorders among HCWs has been associated with negative clinical implications such as poor-quality care and medical errors. Reference Bouaddi, Abdallahi, Fadel Abdi, Hassouni, Jallal and Benjelloun1,Reference Mohammadi, Neyazi, Rangelova, Padhi, Odey and Ogbodum2
Several sociodemographic, health-related, COVID-19-related and work-related correlates of depression, anxiety and PTSD were identified in this review, but are largely limited to one study, making it difficult to draw any conclusions. Nevertheless, there seemed to be consensus in more than one study regarding female gender, younger age, being unmarried, burnout, having a chronic illness, having a history of mental illness, substance use and working at the front line during the COVID-19 pandemic as significant correlates of depressive symptoms among HCWs from SSA. Similarly, for anxiety symptoms, older age, female gender, being married, higher education, lower income, having a chronic illness, neuroticism, resilience, social support and working at the front line during the pandemic were the only correlates with consensus across studies. Female gender, substance use, depression, working at the front line during the pandemic and social support were the only correlates of PTSD with consensus across studies. Many of these factors have also been reported in other reviews. Reference Oyat, Oloya, Atim, Ikoona, Aloyo and Kitara17,Reference De Kock, Latham, Leslie, Grindle, Munoz and Ellis106,Reference Muller, Hafstad, Himmels, Smedslund, Flottorp and Stensland107,Reference Sanghera, Pattani, Hashmi, Varley, Cheruvu and Bradley109,Reference Vizheh, Qorbani, Arzaghi, Muhidin, Javanmard and Esmaeili110 However, since most of these reviews focused on studies conducted during the pandemic, most of the identified factors were COVID-19 related. Empirical studies from outside SSA that were conducted prior to the pandemic also report most of these factors. Reference Cheng and Cheng111–Reference Welsh116
From this review, it appears that the COVID-19 pandemic increased the burden of depression, anxiety and PTSD among HCWs from SSA. Although almost all the included studies did not directly compare the burden of these disorders before and during the pandemic, when comparing the evidence from studies conducted before and during the pandemic the reported prevalence estimates of these disorders across studies appear to be higher during the pandemic compared with before. Evidence from the only study that compared the burden of these, specifically depressive symptoms, among the same set of HCWs before and during the pandemic also shows a significantly higher burden during the pandemic. Reference Quadri, Sultan, Ali, Yousif, Moussa and Fawzy Abdo26 The pandemic also increased the risk of these disorders in this population, with many COVID-19-related risk factors of depression, anxiety and PTSD being identified in this review. Pandemics and other health emergencies have been known to negatively impact the mental health of HCWs. Reference Naushad, Bierens, Nishan, Firjeeth, Mohammad and Maliyakkal11,Reference Stuijfzand, Deforges, Sandoz, Sajin, Jaques and Elmers12,Reference Søvold, Naslund, Kousoulis, Saxena, Qoronfleh and Grobler18 However, in this review, all the evidence is from cross-sectional studies. In the event of similar health emergencies in the future, longitudinal studies are needed not only to explore the immediate impact of such emergencies on healthcare workers’ mental health but also the long-term effects.
Overall, this review shows that a high proportion of HCWs from SSA experience significant symptoms of depression, anxiety and PTSD. These symptoms appear to have been elevated by the COVID-19 pandemic and are associated with several sociodemographic, health-related, COVID-19-related and work-related factors. Targeted interventions are needed to address the burden of common mental disorders in this population.
Strengths and limitations of this review
Our database search included African-based databases, including the African Index Medicus and African Journals Online. This allowed for the inclusion of relevant articles that would potentially have been missed had we focused on only Western-based databases. Additionally, our review was not restricted to a specific time period such as the pandemic, allowing us to provide a comprehensive picture of the mental health of HCWs from this region while also highlighting the potential impact of the COVID-19 pandemic. This review also has certain limitations worth highlighting. Because we restricted our search to only those studies published in English, we might have missed out on other relevant studies published in other languages. Additionally, all the included studies were cross-sectional in design, limiting our conclusions about causality and temporal associations. The limited information regarding the reliability and validity of the measurement tools used in most studies, as well as the non-random recruitment of study participants, may have introduced bias that undermines the accuracy, conclusions and generalisability of study findings.
Implications of the results for practice, policy and future research
Despite the limitations, this study has implications for practice, policy and future research. The high burden of depression, anxiety and PTSD among HCWs from SSA calls for an urgent need for policies and interventions to address this burden in this key population. Such interventions should be multi-component in nature, given the multifactorial nature of the correlates of these disorders. The high burden also calls for the integration of mental health screening and treatment into the available healthcare package for HCWs in SSA. All the included studies were cross-sectional in design. There is a need for future research to focus on longitudinal designs to map the longitudinal trends of mental health among HCWs. Future research could also utilise different study designs to explore causality and confidently ascertain the reported associations. The heterogeneity of the outcome measures used precluded a meta-analysis. We recommend that future studies use standardised measures to allow for accurate and meaningful comparisons. Finally, since very few of the studies had recruited comparison groups, future studies on the mental health of HCWs should aim to include comparison groups to enable more definite conclusions.
Supplementary material
The supplementary material is available online at https://doi.org/10.1192/bjo.2025.10818
Data availability
The data supporting the conclusions presented in this article are available within this article and its supplementary material.
Acknowledgement
The authors thank Gideon Mbithi for helping develop the map showing the geographic distribution of the included studies.
Author contributions
Conceptualisation, A.A., E.N.-M. and S.A.O. Methodology, A.A., E.N.-M., E.K.T., S.A.O., P.W. and S.N. Data extraction, P.W., S.N. and E.K.T. Quality assessment, P.W., S.N. and E.K.T. Data synthesis, E.K.T. Writing (original draft preparation), E.K.T. Writing (review and editing), S.A.O., P.W., S.N., E.N.-M. and A.A. Funding acquisition, A.A. and E.N.-M. All authors read, provided feedback on and approved the submitted version of the manuscript.
Funding
This work was funded by the Johnson and Johnson Foundation (grant no. 63773339). During this research, E.K.T. and A.A. were supported by a grant from Science for Africa Foundation (grant no. Del-22-002) with funding from the Wellcome Trust and the UK Foreign, Commonwealth and Development Office, which is part of the EDCTP 2 programme supported by the European Union. A.A.’s work is also supported by the Office of the Director, National Institutes of Health (OD), the National Institute of Biomedical Imaging and Bioengineering (NIBIB), the National Institute of Mental Health (NIMH) and the Fogarty International Center (FIC) of the National Institutes of Health, under award no. U54TW012089 (A.A. and A.K. Waljee). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Johnson and Johnson Foundation. The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript. No additional external funding was received for this study.
Declaration of interest
None.
Open access
eLetters
No eLetters have been published for this article.