Abstract
Introduction
Depression and post-traumatic stress disorder (PTSD) significantly contribute to the global disease burden, adversely affecting health, medical costs and economic productivity. The SMART-DAPPER project in western Kenya approached this challenge by deploying trained non-specialists to administer proven depression and PTSD treatments.
Methods
Participants were public sector primary care outpatients at Kiumu County Hospital with major depression and/or PTSD. They were randomised to first-line treatment with interpersonal psychotherapy (IPT) or fluoxetine. We evaluated three measures of economic productivity: income, absenteeism and presenteeism (lowered work efficiency). Change over baseline and comparisons between treatments were conducted using generalised estimating equations regression.
Results
There were statistically significant gains in economic productivity from baseline to the end of first-line treatment. The percentage of participants earning a monthly income rose from 54.9% at baseline to 59.8% after treatment in the IPT group and from 54.5% to 61.5% in the Fluoxetine group. Improvement was significantly associated with illness remission. Average monthly income among earners increased by Kenya shillings (KES) 1920 (46 Intl$) with IPT and KES 1350 (31 Intl$) with fluoxetine. Absenteeism dropped in both treatment arms, by 1.5 days per month for IPT and 1.9 days per month for fluoxetine. Presenteeism decreased more with fluoxetine (4.8 days per month) than with IPT (3.3 days per month).
Conclusion
Treating common mental disorders with IPT and fluoxetine in public sector primary care settings was associated with economic productivity. Leveraging a non-specialist workforce for treatment delivery at scale may build individual and community economic well-being.
Funding
R01MH113722(NIMH), R01MH115512(NIMH-GACD).
Trial registration number
ClinicalTrials.gov Identifier: NCT03466346.
Keywords: Global Health, Health economics, Health policy, Mental Health & Psychiatry, Kenya
WHAT IS ALREADY KNOWN ON THIS TOPIC.
WHAT THIS STUDY ADDS
In this randomised clinical trial of 2162 adults, the percentage of participants earning a monthly income rose from 54.9% at baseline to 59.8% after treatment in the interpersonal psychotherapy group and from 54.5% to 61.5% in the fluoxetine group. Improvement was significantly associated with illness remission. Income levels also rose. There were declines in absenteeism and presenteeism (lower work efficiency).
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Evidence-based psychotherapy and medication for depression and PTSD can be delivered by local non-specialist clinicians to public sector primary care patients in a low-income African setting and is associated with gains in economic productivity.
Introduction
According to global health surveys by the WHO, mental disorders account for a huge proportion of the global burden of disability due to disease (14%), and the most common mental disorder (depression) is the single leading cause of global disability.1 The prevalence of mental disorders and associated disability is particularly high in Sub-Saharan Africa.2 This morbidity is compounded by a small mental health workforce and scarce access to evidence-based, curative mental healthcare—the ‘mental health treatment gap’.
A high prevalence of mental disorders exacts an extraordinary economic toll globally, including increased overall healthcare costs as well as decreased economic productivity affecting households and communities. Patients with depression often present to healthcare systems with clinical manifestations such as headache, stomach ache, dizziness, fatigue or pain.3 4 Without a robust mental health workforce, these symptoms are typically evaluated as physical problems with associated costs, such as blood tests and irrational prescriptions (eg, overuse of antibiotics). When these treatments fail, patients may return or seek care elsewhere, using scarce public sector health worker time and further spurious and expensive tests/treatment. Depression compromises economic productivity, typically affecting individuals during years of peak productivity (starting in early adulthood).5 Without adequate treatment, depression may persist across decades, leading to ‘absenteeism’ and ‘presenteeism’. Absenteeism is the absence of a worker (eg, low energy or lack of interest, causing the workers to stay home), while presenteeism refers to reduced efficiency at work (eg, poor concentration).6,8 Studies in high-income countries (HICs) find that employees with depression lose 20% of total work effort, of which 81% are due to presenteeism and 19% to absenteeism.9,11 Treatment for depression is associated with a 300% rate of return in HICs, when comparing the cost of treatment with decreased healthcare costs and improved economic productivity.12,16
The SMART-DAPPER project in western Kenya is a Sequential, Multiple Assignment Randomized Trial (SMART) that trained and deployed a readily available, non-specialist workforce to provide established treatments for the two most common mental disorders in the region (depression and post-traumatic stress disorder (PTSD)) for adult public sector primary care patients.17 The trial used two evidence-based first-line treatments for depression and PTSD: fluoxetine, a selective serotonin reuptake inhibitor (SSRI), delivered by nurses and clinical officers (authorised to prescribe) with no previous training in mental health, and interpersonal psychotherapy (IPT), delivered by non-specialised health and lay providers. Fluoxetine is on the WHO list of essential medications for Kenya and is available nationwide to public sector healthcare facilities through the Kenya Medical Supplies Authority. IPT has been previously adapted and tested in the region and shows strong efficacy.18 During the initial part of the study, all treatment sessions were in person, but from 2020 during the pandemic, patients had a choice of in-person or telephone (see study protocol for details).17
The SMART DAPPER trial employed robust economic measures to evaluate the economic benefits of treatment at the participant level. This information is key to establishing the case for increasing access to mental health services in the health policy agenda and in the initiatives to improve labour productivity. These economic data can inform the Kenyan government’s current efforts to develop robust mental health service scale-up plans and budgeting. The Kenya Mental Health Policy 2015–203019 recommends efficient management of human resources for mental healthcare provision through continuous education and professional development, equitable deployment and motivation to retain service providers at all healthcare levels including supportive supervision and coordination. In response, policymakers and healthcare providers have launched a government-funded initiative to scale up treatment for mental disorders in primary healthcare.19
In the analysis reported here, we assess economic productivity outcomes associated with the first round of treatment (with a remission rate of 82%), before the treatment reassignment phase of the trial.
Methods
Setting and participants
Data were collected between September 2020 and July 2022, from baseline (pretreatment) through month 18. Participants were recruited from the general outpatient primary care services at the Kisumu County Referral Hospital and other nearby public health facilities and were included if they screened positive for major depressive disorder (MDD) and/or PTSD on the Mini-International Neuropsychiatric Interview (MINI), were 18 years of age or older and could attend weekly IPT sessions/fluoxetine monitoring appointments. Exclusion criteria included pregnancy or breastfeeding, history of mania or hypomania and cognitive dysfunction limiting psychotherapy participation. Those with active suicidality, drug/alcohol use disorders requiring substance use treatment or current hypomania/mania were excluded and referred for higher level or appropriate mental health treatment.
Study measures
The study used a questionnaire adapted from the World Bank Living Standards Measurement Study (LSMS).20 LSMS designed surveys to collect data on many economic and health dimensions of household well-being, including consumption, income, savings, employment, health, education, fertility, nutrition, housing and migration. For all our analyses, outcomes were examined as repeated measures: baseline and follow-up at 3, 6 and 9 months. End of treatment was included in the analysis even though the duration of treatment differed. We calculated the economic benefits to individuals associated with treatment, both separately for IPT and fluoxetine and pooled.
Specific productivity outcomes were the proportion of participants receiving a monthly income, mean income earned, and the number of days of absenteeism (work hours lost due to illness or doctor visits) and presenteeism (days in the past month partially unable to work). Income was based on wages from work done for a non-household member, agricultural jobs, business and self-employment in the past month. Productivity gains in the informal and agricultural sectors were converted to monetary values using net income in the local labour market. All currency measurements were in Kenya shillings (KES), converted to international dollars (Int$) based on World Bank purchasing power parity for 2021 of 43.33 KES per Int$.21
Statistical analysis
We described our data using proportions (percentages) for receiving an income and means for income, absenteeism and presenteeism. Differences from baseline to follow-up and across treatment arms as well as differences between those who did or did not achieve clinical remission at the end of treatment were tested for statistical significance using generalised estimating equations regressions with robust standard errors; logistic for proportions (quantifying effects as ORs); and linear for continuous measures (quantifying effects as differences in means) as well as 95% CIs. Comparisons between treatment arms incorporated interactions of time (baseline vs end of follow-up) and treatment arm; remission and time status; and 3-way time, remission status and intervention arm, depending on the question addressed.
Ethics
The trial was approved by the UCSF Institutional Review Board and the Kenyatta National Hospital-University of Nairobi Ethics and Research Committee and the Kenya Pharmacy and Poisons Board. It was registered on ClinicalTrials.gov (NCT03466346) and conducted by Human Subjects Protections and Good Clinical Practice. The trial was monitored by the United States National Institutes for Health through PPD.
Results
The demographic and clinical characteristics of the participants at baseline were similar across the two treatment arms (table 1). The average age of participants was 36 years, and the majority were married women with some primary or secondary education. At baseline, over 90% of participants had a diagnosis of major depression; approximately 52% had PTSD. Nearly half had both major depression and PTSD.
Table 1. Baseline characteristics by treatment arm.
| IPT (n=1082) | Fluoxetine (n=1080) | Total (n=2162) | |
|---|---|---|---|
| Age | |||
| Age in years (mean±SD (n)) | 36.1±11.2 (n=1082) | 35.3±10.8 (n=1080) | 35.7±11 (n=2162) |
| Age in years (median (min-max)) | 34 (18–85) | 33 (18–77) | 34 (18–85) |
| Gender | |||
| Female | 977 (90.3%) | 981 (90.8%) | 1958 (90.6%) |
| Male | 105 (9.7%) | 99 (9.2%) | 204 (9.4%) |
| Formal Education | |||
| None | 16 (1.5%) | 19 (1.8%) | 35 (1.6%) |
| Some primary/primary | 576 (53.2%) | 541 (50.1%) | 1117 (51.7%) |
| Some secondary/secondary | 389 (36.0%) | 418 (38.7%) | 807 (37.3%) |
| Some college/certificate/diploma | 101 (9.3%) | 102 (9.4%) | 203 (9.4%) |
| Baseline diagnosis(es) (MINI) | |||
| Major depression | 1009 (93.3%) | 1006 (93.1%) | 2015 (93.2%) |
| PTSD | 555 (51.3%) | 563 (52.1%) | 1118 (51.7%) |
| Major depression and PTSD | 498 (46.0%) | 508 (47.0%) | 1006 (46.5%) |
| Depression symptoms (BDI) (mean±SD (n)) | 28.7±10.3 (n=1082) | 29.1±10.5 (n=1079) | 28.9±10.4 (n=2161) |
| PTSD symptoms (PCL) (mean±SD (n)) | 43.6±17.4 (n=1082) | 43.3±17.2 (n=1079) | 43.5±17.3 (n=2161) |
| History mental healthcare | 10 (0.9%) | 13 (1.2%) | 23 (1.1%) |
| Co-morbidities | |||
| HIV | 438 (40.5%) | 413 (38.2%) | 851 (39.4%) |
| Other* | 99 (9.1%) | 96 (8.9%) | 195 (9.0%) |
| Trauma | |||
| Physical intimate partner violence in past week among partnered participants (CTS)** | 319 (57.9%) | 343 (60.5%) | 662 (59.2%) |
| No lifetime trauma history (THQ) | 93 (8.6%) | 80 (7.4%) | 173 (8.0%) |
| One type of lifetime trauma (TH)Q | 289 (26.7%) | 300 (27.8%) | 589 (27.3%) |
| Two types of lifetime trauma (THQ) | 453 (41.9%) | 479 (44.4%) | 932 (43.1%) |
| Three or more types of lifetime trauma (THQ) | 247 (22.8%) | 220 (20.4%) | 467 (21.6%) |
IPT, interpersonal psychotherapy; MINI, Mini-International Neuropsychiatric Interview; PTSD, post-traumatic stress disorder.
Both treatments were associated with significant increases in the percentage of participants earning a monthly income (table 2). This percentage rose during the first line treatment from 54.9% at baseline to 59.8% for IPT and from 54.5% to 61.5% for fluoxetine. These represented gains of OR=1.22 (95% CI 1.06 to 1.40) and 1.34 (1.15 to 1.56), respectively. The average monthly income among earners increased significantly by KES 1920 (816, 3057; 45.96 Intl$) and KES 1350 (837, 1893; 31.18 Intl$) in the IPT and fluoxetine groups. At the end of the first line treatment, there was a significant reduction of absenteeism for IPT −1.5 (−1.8 to –1.1) days and fluoxetine −1.9 (−2.3 to –1.5) days per month. There were no statistically significant differences across treatment arms for these outcomes. Days of presenteeism (partial inability to work) decreased significantly more with fluoxetine −4.8 (–5.4 to –4.2) than with IPT −3.3 (–3.9 to –2.7), a significant difference by arm of −0.73 (–1.2 to –0.26) when adjusted for baseline.
Table 2. Economic outcomes: end of first-line treatment.
| Treatment arm | Baseline | End of treatment | Difference between end of treatment and baseline (OR or difference, CI, p value) | Between group comparison at end of treatment (OR, CI, p value) | |
|---|---|---|---|---|---|
| Participants that earned a monthly income (%) – y/n | IPT | 594 (54.9%) | 639 (59.8%) | 1.22 (1.06, 1.40) p=0.0060 | |
| FLX | 589 (54.5%) | 629 (61.5%) | 1.34 (1.15, 1.56) p=0.0002 | 1.09 (0.91, 1.31) p=0.35 | |
| Average monthly income among income earners (Kenyan shillings) | IPT | 4190±5490 (n=594) | 6110±13 500 (n=639) | 1936 (816, 3057) p=0.0007 | |
| FLX | 3900±5450 (n=589) | 5250±5070 (n=629) | 1364 (837, 1893) p=<0.0001 | −885 (−2468, 698) p=0.27 | |
| Absenteeism: days in the past month completely unable to work (mean #) | IPT | 2.5±4.8 (n=1082) | 1.0±3.3 (n=986) | −1.5 (−1.8, −1.1) p=<0.0001 | |
| FLX | 2.7±5.3 (n=1080) | 0.8±2.9 (n=933) | −1.9 (−2.3, −1.5) p=<0.0001 | −0.23 (−0.51, 0.046) p=0.10 | |
| Presentism: days in the past month partially unable to work (mean #) | IPT | 5.3±8.5 (n=1082) | 2.0±5.9 (n=986) | −3.3 (−3.9, −2.7) p=<0.0001 | |
| FLX | 6.1±9.5 (n=1080) | 1.3±4.6 (n=933) | −4.8 (−5.4, −4.2) p=<0.0001 | −0.73 (−1.2, −0.26) p=0.0024 |
ORs are reported for the dichotomous outcome (earning a monthly income) and means for continuous outcomes (monthly income and days of absenteeism and presenteeism). In the first row, the statistical test in the final column is based on the ratio of ORs from prior columns, adjusted for baseline values. In the subsequent rows, the tests are based on differences in means, adjusted for baseline values.
FLX, Fluoxetine; IPT, interpersonal psychotherapy.
Some economic effects differed between clinical remitters and non-remitters at the end of first-line treatment (table 3). The increase in proportion earning a monthly income was higher for IPT remitters (11.2%) than non-remitters (2.0%) (p=0.03). For Fluoxetine (FLX), the difference was proportionally smaller (13.5% vs 6.7%) and not statistically significant. For monthly income among earners, remitters did not differ significantly from non-remitters. For absenteeism and presentism days, IPT remitters had a larger drop than non-remitters (p=0.01 and 0.04, respectively); FLX did not. Remission effects did not differ significantly across treatment arms.
Table 3. Economic outcomes: remitters and non-remitters at end of first-line treatment.
| For remitters at end of first-line treatment | For non-remitters at end of first-line treatment | P-values for interactions | |||||
|---|---|---|---|---|---|---|---|
| Baseline | Follow-up: end of first line | Baseline | Follow-up: end of first line | Time by remitter status | Time by remission by arm | ||
| Participants that earned a monthly income (%) | IPT | 439 (56.1%) | 527 (67.3%) | 108 (53.2%) | 112 (55.2%) | 0.03 | n/a |
| FLX | 441 (55.2%) | 549 (68.7%) | 71 (53.0%) | 80 (59.7%) | 0.16 | 0.74 | |
| Average monthly income among income earners (Kenyan shillings) | IPT | 4320±5940 (n=439) | 6280±14 400 (n=527) | 3440±3040 (n=108) | 5300±7910 (n=112) | 0.99 | n/a |
| FLX | 3800±4610(n=441) | 5430±4990 (n=549) | 3360±3000 (n=71) | 4010±5450 (n=80) | 0.08 | 0.37 | |
| Absenteeism: days in the past month completely unable to work (mean #) | IPT | 2.3±4.6 (n=783) | 0.6±2.3 (n=783) | 3.0±5.5 (n=203) | 2.61±5.27 (n=203) | 0.01 | n/a |
| FLX | 2.4±5.0 (n=799) | 0.4±2.0 (n=799) | 4.2±6.7 (n=134) | 2.84±5.5 (n=134) | 0.45 | 0.41 | |
| Presentism: days in the past month partially unable to work (mean #) | IPT | 5.2±8.5 (n=783) | 1.4±5.0 (n=784) | 6.5±9.5 (n=203) | 4.6±7.8 (n=203) | 0.04 | n/a |
| FLX | 5.7±9.1 (n=799) | 0.8±3.7 (n=799) | 8.5±10.6 (n=134) | 4.4±7.5 (n=134) | 0.53 | 0.48 | |
Percentages are reported for the dichotomous outcome (earning a monthly income) and means for continuous outcomes (monthly income and days of absenteeism and presenteeism). In the final column, the statistical test assesses differencs across treatment arm.
Several productivity measures among remitters improved over time following the end of treatment (table 4 and figure 1). Remitters who had received IPT had higher odds of earning a monthly income at 6 months (OR 1.29, 1.09–1.53) and 9–12 months (OR 1.24, 1.04–1.48) than at the end of treatment. Remitters in the FLX arm had a higher monthly income at 9–12 months than at the end of treatment. There was no evolution of gains in days of absenteeism. Presentism dropped in the IPT arm at 9–12 months, compared with the end of treatment: −0.5 days (−0.9, 0.0).
Table 4. Economic outcomes: end of first-line treatment through 12 months after first-line treatment completion among remitters: evolution of gains.
| TRT | End of TRT | Time after TRT completion | ORs or means between end of TRT and time after TRT (95% CI), p value | Ratios of OR or means between TRT arms (95% CI), p value | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 3 mo | 6 mo | 9–12 mo | 3 mo | 6 mo | 9–12 mo | 3 mo | 6 mo | 9–12 mo | |||
| Participants that earned a monthly income (%) | IPT | 527 (67.3%) | 523 (67.1%) | 548 (73.1%) | 542 (72.3%) | 0.99 (0.83, 1.17) p=0.87 | 1.29 (1.09, 1.53) p=0.0036 | 1.24 (1.04, 1.48) p=0.018 | |||
| FLX | 549 (68.7%) | 519 (68.4%) | 539 (71.4%) | 540 (72.1%) | 0.97 (0.82, 1.15) p=0.75 | 1.11 (0.93, 1.33) p=0.25 | 1.16 (0.96, 1.40) p=0.12 | 0.99 (0.78, 1.25) p=0.91 | 0.86 (0.67, 1.10) p=0.24 | 0.94 (0.72, 1.22) p=0.63 | |
| Average monthly income among income earners (Kenyan shillings) | IPT | 6280±14 400 (n=527) | 6530±7910 (n=523) | 6310±7980 (n=548) | 7070±11 500 (n=542) | 222 (−1161, 1604) p=0.75 |
76 (−1249, 1402) p=0.91 |
866 (−471, 2203) p=0.20 |
|||
| FLX | 5430±4990 (n=549) | 5930±5610 (n=519) | 5920±6150 (n=539) | 6520±6580 (n=540) | 429 (−37, 896) p=0.071 |
491 (−59, 1042) p=0.08 |
1147 (564, 1731) p=0.0001 |
207 (−1252, 1666) p=0.78 |
415 (−1020, 1850) p=0.57 |
281 (−1177, 1740) p=0.71 |
|
| Absenteeism: days in the past month completely unable to work (mean #) | IPT | 0.6±2.3 (n=783) |
0.7±2.7 (n=743) |
0.5±2.2 (n=745) | 0.4±1.8 (n=747) | 0.1 (−0.1, 0.4) p=0.3291 | −0.1 (−0.3, 0.2) p=0.5910 | −0.2 (−0.4, 0.0) p=0.0754 | |||
| FLX | 0.4±2.0 (n=799) |
0.5±1.9 (n=753) |
0.7±3.1 (n=745) | 0.5±2.5 (n=738) | 0.0 (-0.2, 0.2) p=0.89 | 0.2 (0.0, 0.5) p=0.053 | 0.1 (−0.2, 0.3) p=0.65 | −0.1 (−0.4, 0.20) p=0.48 | 0.3 (0.0, 0.6) p=0.077 | 0.2 (−0.1, 0.5) p=0.12 | |
| Presentism: days in the past month partially unable to work (mean #) | IPT | 1.4±5.0 (n=784) |
1.3±4.5 (n=744) |
1.1±4.4 (n=745) | 0.9±3.8 (n=747) | 0.0 (−0.5, 0.4) p=0.88 | −0.3 (−0.7, 0.2) p=0.20 | −0.5 (−0.9, 0.0) p=0.037 | |||
| FLX | 0.8±3.7 (n=799) | 1.1±4.0 (n=753) |
0.8±3.5 (n=745) | 0.8±3.5 (n=738) | 0.2 (−0.2, 0.6) p=0.26 | 0.0 (−0.4, 0.4) p=0.99 | 0.0 (−0.4, 0.3) p=0.86 | 0.3 (−0.4, 0.9) p=0.40 | 0.3 (−0.3, 0.8) p=0.33 | 0.4 (−0.1, 1.0) p=0.13 | |
Percentages and ORs are reported for the dichotomous outcome (earning a monthly income), and means for continuous outcomes (monthly income, and days of absenteeism and presenteeism). Statistical significance in the middle block assesses change over time within arm (with differences in means or ORs), and in the third block, difference-of-differences in means across treatment arms or ratios of ORs. Significant p-values are bolded.
P is the significance level.
FLX, Fluoxetine; IPT, interpersonal psychotherapy; TRT, Treatment used.
Figure 1. Monthly income (Kenyan shillings) among income earners among remitters. FLX, Fluoxetine; IPT, interpersonal psychotherapy.
Interpretation
These SMART DAPPER findings suggest that evidence-based treatment of major depression and PTSD delivered by a readily available non-specialist workforce to adult public sector primary care patients is associated with improved economic productivity: increased wage income and decreased absenteeism and presenteeism. Some of these gains correlated with clinical remission. The mental health service implementation strategy used by SMART DAPPER may provide local and national policymakers with valuable information not only to alleviate human suffering but also to improve the economic status of individuals, households and communities. This should provide added resources necessary to advance society (e.g., education costs).
The study’s findings provide valuable insights into the potential effects of IPT and FLX treatments on income-related outcomes and work productivity. FLX and other SSRIs have been shown to improve overall functioning and reduce work-related impairments in individuals with depression, which may explain the observed reduction in presentism with FLX treatment. The findings suggest that IPT may offer advantages in terms of income earnings compared with FLX over an extended period post-treatment. Additionally, reduced workplace productivity and functioning associated with MDD imposes a significant economic burden. Our study findings reported improvements in productivity following both IPT and FLX treatments. The finding that FLX had a stronger association with reduced presentism than did IPT aligns with findings in the neighbouring country of Uganda, where anti-depressant treatment among people living with HIV in Uganda was linked with improved work functioning.22 Similarly, a randomised trial evaluating enhanced depression care for primary care patients in Harvard reported an 8% increase in productivity among consistently employed workers over 2 years.23 A recent systematic review found that anti-depressant treatment generally improves measures of workplace functioning and productivity in patients with MDD.24 Similarly, the results of a pharmacoeconomic analysis reported that the observed improvements in cognitive symptoms and workplace productivity in working patients with MDD treated with fluoxetine resulted in a mean financial gain for the employer of CAN$110.64 per patient per week.25 Regarding absenteeism, both IPT and FLX groups demonstrated a reduction in the number of days missed over time, showing improvements from baseline to the end of treatment.
Potential significance to employers
The productivity gains reported in this study represent significant gains for employers. Thus, the findings underscore the benefit to the government and public sector of investing in treatment for common mental disorders. Notably, presenteeism (inefficiency at work due to a disorder) should be addressed to ensure employees get the required treatment in time. Presenteeism can be an antecedent to absenteeism (absence from work) and disability leave. Employees are limited in what they can accomplish due to their depressive symptoms. Ultimately, when absent from work, an employee is likely incapacitated by symptoms of depression (eg, loss of interest or fatigue) or seeking healthcare services. Hence, it is crucial to identify common mental disorders, such as depression, early in their course and refer for evidence-based treatment in order to prevent future long-term mental health disability costs to employers. Similarly, a previous cohort study reported that generally ill people with no absenteeism had two times the risk of heart attack than those with some absenteeism.26
Role in universal health insurance
In Kenya, the inclusion of IPT and FLX in Universal Health Coverage (UHC) would be particularly valuable when integrated within social health insurance systems. Integration of IPT and FLX into UHC programmes ensures that individuals presenting with depression and/or PTSD receive evidence-based treatment, boosting their well-being and productivity. This integration is vital for fostering sustainable economic growth, especially within the context of social health insurance in Kenya. The benefit package for social health insurance includes coverage for mental health services such as mental health education, counselling, screening, bio/psychosocial treatment and referral when indicated for various disorders including behavioural, neurodevelopmental, affective, psychotic and psychoactive use disorders. Additionally, rehabilitation services for psychoactive use disorders and behavioural addictions are provided, demonstrating a comprehensive approach to mental healthcare within the social health insurance framework in Kenya.
Limitations
Some limitations of this study should be noted. First, the emphasis on income and employment might be interpreted as undermining the importance of treating common mental disorders among groups who are not in the workforce, such as the elderly. To the contrary, we believe that all individuals who suffer from mental disorders should be treated and note that productivity gains in the informal sector and improved functioning bring important economic benefits (which we will report separately). Second, data on time lost from work may be subject to recall bias and imprecision. Third, we lacked detailed data on other medical or psychiatric conditions that might be associated with work loss and productivity impairment. Fourth, this dataset does not include the study’s longer-term follow-up assessments; this will be published when available and may reveal differences in economic productivity according to treatment assignment that are not yet apparent.
Fifth, this trial compared two treatment approaches, without a non-intervention control arm, because our purpose was to assess comparative effectiveness in a real-world setting with delivery by primary care integrated non-specialists. Thus, we cannot precisely quantify the treatment effect. However, data from high-income countries establish that short-term spontaneous remission of depression is uncommon, occurring in only about 12% of individuals within 12 weeks (IPT time frame in this study).27 In addition, this project was based on our previous work in the region including an RCT in which IPT recipients had greater remission vs wait-list controls (risk ratio 0.64).18
Similarly, lack of a non-intervention control group means that the evidence of improved economic productivity should be treated as descriptive, an association rather than direct evidence of causality. Higher productivity with remission as a post-randomisation variable (eg, tables3 4) is suggestive but not definitive for causality. Also, non-remitters had worse productivity at baseline; this association with slow clinical response is unsurprising but suggests cautious interpretation.
Relatedly, our study did not measure background secular economic trends during the course of SMART DAPPER that might partially account for the increase in percentage of people making a monthly income and the amount of that income.
Finally, we highlight the predominance of female participants. Gender disparities have a significant impact on disease burden, including mental conditions and are believed to be one of the reasons for the higher prevalence of common mental disorders such as depression and trauma-related conditions among females.28 29 Furthermore, females are more likely to seek healthcare, including mental healthcare, than males.30 31 The vast majority of the participants in the SMART DAPPER project were women, consistent with global expectations for disease burden and healthcare seeking behaviours. In this region of Kenya which has a high prevalence of HIV, the gender disparity in primary care might be amplified by national policies that require HIV testing for pregnant women, which could lead to more women than men being aware of their HIV status and enrolled in care.
Conclusions
The findings from SMART DAPPER support the economic case for investing in a scalable strategy using non-specialists to deliver IPT and fluoxetine for depression and PTSD treatment to public sector primary care patients in the region. Further analysis is required to estimate the return on investments and the cost-benefit of IPT versus fluoxetine for the treatment of common mental disorders in East African primary care settings and to promote understanding of positive implications for employees, employers, their families and society at large.
Acknowledgements
We thank our participants. We are grateful for the privilege of partnering with the Kenyan National and County Ministries of Health in the development of mental health care services in the region. Data Sharing. Our data will be publicly available through the National Institute of Mental Health Data Archive (NDA), an online repository for human subjects’ data.
Footnotes
Funding: R01MH113722(NIMH), R01MH115512(NIMH-GACD). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Handling editor: Lei Si
Data availability free text: Data are available in the National Institute of Mental Health Data Archive.
Patient consent for publication: Consent obtained directly from patient(s).
Ethics approval: This study involved human participants and was approved by the UCSF, Human Research Protection Program Institutional Review Board (IRB) (18-24852), Kenyatta National Hospital and University of Nairobi Ethics and Research Committee (P344/05/2018). Participants gave informed consent to participate in the study before taking part.
Provenance and peer review: Not commissioned; externally peer-reviewed.
Patient and public involvement: Patients and members of the public were involved in several stages of the trial, including the design, management and conduct of the trial. SMART DAPPER utilized the Exploration, Preparation, Implementation, Sustainment (EPIS) implementation framework. The model guided the engagement of Implementation Resource Team (IRT) stakeholders from the inner context (e.g., clinic staff, patients and providers) and the outer context (e.g., community leaders and county, national and regional health policy leaders). Study leaders met with the IRT prior to the study launch and throughout the study to foster collaboration, incorporate stakeholder voices, and monitor and tailor the intervention for population and contextual needs, including incorporating mobile phone treatment. We meet annually with the IRT to disseminate study findings and seek patient and public involvement in the development of an appropriate method of dissemination.
Data availability statement
Data are available in a public, open access repository.
References
- 1.Mental health of adolescents. 2024. https://www.who.int/news-room/fact-sheets/detail/adolescent-mental-health Available.
- 2.Ferrari A. Global, regional, and national burden of 12 mental disorders in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry. 2022;9:137–50. doi: 10.1016/S2215-0366(21)00395-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wise TN, Fishbain DA, Holder-Perkins V. Painful Physical Symptoms in Depression: A Clinical Challenge. Pain Med. 2007;8:S75–82. doi: 10.1111/j.1526-4637.2007.00352.x. [DOI] [PubMed] [Google Scholar]
- 4.Frequency and pattern of common primary headache among depressed patients at tertiary care centre, Karachi. 2024. https://pubmed.ncbi.nlm.nih.gov/29171561/ Available. [PubMed]
- 5.de Oliveira C, Saka M, Bone L, et al. The Role of Mental Health on Workplace Productivity: A Critical Review of the Literature. Appl Health Econ Health Policy. 2023;21:167–93. doi: 10.1007/s40258-022-00761-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Howard KJ, Howard JT, Smyth AF. Handbook of occupational health and wellness. 2012. The problem of absenteeism and presenteeism in the workplace; pp. 151–79.https://link.springer.com/chapter/10.1007/978-1-4614- 4839-6_8 Available. [Google Scholar]
- 7.Lamichhane DK, Heo YS, Kim HC. Depressive symptoms and risk of absence among workers in a manufacturing company: a 12-month follow-up study. Ind Health. 2018;56:187–97. doi: 10.2486/indhealth.2017-0065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Johnston DA, Harvey SB, Glozier N, et al. The relationship between depression symptoms, absenteeism and presenteeism. J Affect Disord. 2019;256:536–40. doi: 10.1016/j.jad.2019.06.041. [DOI] [PubMed] [Google Scholar]
- 9.Gómez-Restrepo C, Bohórquez A, Pinto Masis D, et al. Prevalencia de depresión y factores asociados con ella en la población colombiana Investigación original / Original research Los resultados del estudio sobre la carga mundial de las enfermedades (Global Burden of Disease, GBD) efec-tuado por la Organización Mundial de Palabras clave. Rev Panam Salud Publica/Pan Am J. 2004;16:378–86. doi: 10.1590/S1020-49892004001200003. [DOI] [PubMed] [Google Scholar]
- 10.Gudleski GD, Satchidanand N, Dunlap LJ, et al. Predictors of medical and mental health care use in patients with irritable bowel syndrome in the United States. Behav Res Ther. 2017;88:65–75. doi: 10.1016/j.brat.2016.07.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gómez-Restrepo C, Bohórquez A, Pinto Masis D, et al. Prevalencia de depresión y factores asociados con ella en la población colombiana. Rev Panam Salud Publica. 16:378–86. doi: 10.1590/S1020-49892004001200003. n.d. [DOI] [PubMed] [Google Scholar]
- 12.Clark M, DiBenedetti D, Perez V. Cognitive dysfunction and work productivity in major depressive disorder. Expert Review of Pharmacoeconomics & Outcomes Research. 2016;16:455–63. doi: 10.1080/14737167.2016.1195688. [DOI] [PubMed] [Google Scholar]
- 13.Lachaine J, Beauchemin C, Bibeau J, et al. Canadian economic impact of improved workplace productivity in patients with major depressive disorder treated with vortioxetine. CNS Spectr. 2020;25:372–9. doi: 10.1017/S1092852919000853. [DOI] [PubMed] [Google Scholar]
- 14.Lerner D, Adler DA, Rogers WH, et al. A Randomized Clinical Trial of a Telephone Depression Intervention to Reduce Employee Presenteeism and Absenteeism. PS. 2015;66:570–7. doi: 10.1176/appi.ps.201400350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang PS, Simon GE, Avorn J, et al. Telephone Screening, Outreach, and Care Management for Depressed Workers and Impact on Clinical and Work Productivity Outcomes. JAMA. 2007;298:1401. doi: 10.1001/jama.298.12.1401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rost K, Smith JL, Dickinson M. The effect of improving primary care depression management on employee absenteeism and productivity. A randomized trial. Med Care. 2004;42:1202–10. doi: 10.1097/00005650-200412000-00007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Levy R, Mathai M, Chatterjee P, et al. Implementation research for public sector mental health care scale-up (SMART-DAPPER): a sequential multiple, assignment randomized trial (SMART) of non-specialist-delivered psychotherapy and/or medication for major depressive disorder and posttraumatic stress disorder (DAPPER) integrated with outpatient care clinics at a county hospital in Kenya. BMC Psychiatry. 2019;19:424. doi: 10.1186/s12888-019-2395-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Meffert SM, Neylan TC, McCulloch CE, et al. Interpersonal psychotherapy delivered by nonspecialists for depression and posttraumatic stress disorder among Kenyan HIV-positive women affected by gender-based violence: Randomized controlled trial. PLoS Med. 2021;18:e1003468. doi: 10.1371/journal.pmed.1003468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.MoH Kenya mental health policy. 2015 https://publications.universalhealth2030.org/uploads/Kenya-Mental-Health-Policy.pdf Available.
- 20.Living standards measurement study. 2024. https://www.worldbank.org/en/programs/lsms Available.
- 21.PPP conversion factor, GDP (LCU per international $) - Kenya | Data. 2024. https://data.worldbank.org/indicator/PA.NUS.PPP?locations=KE Available.
- 22.Wagner GJ, Ghosh-Dastidar B, Slaughter M, et al. The Role of Depression in Work-Related Outcomes of HIV Treatment in Uganda. IntJ Behav Med. 2014;21:946–55. doi: 10.1007/s12529-013-9379-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Rost K, Smith JL, Dickinson M. The Effect of Improving Primary Care Depression Management on Employee Absenteeism and Productivity. Med Care. 2004;42:1202–10. doi: 10.1097/00005650-200412000-00007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lee Y, Rosenblat JD, Lee J, et al. Efficacy of antidepressants on measures of workplace functioning in major depressive disorder: A systematic review. J Affect Disord. 2018;227:406–15. doi: 10.1016/j.jad.2017.11.003. [DOI] [PubMed] [Google Scholar]
- 25.Ongeri L, McCulloch CE, Neylan TC, et al. Suicidality and associated risk factors in outpatients attending a general medical facility in rural Kenya. J Affect Disord. 2018;225:413–21. doi: 10.1016/j.jad.2017.08.059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kimura K, Nagata T, Ohtani M, et al. Cardiovascular and cerebrovascular diseases risk associated with the incidence of presenteeism and the costs of presenteeism. J Occup Health. 2020;62 doi: 10.1002/1348-9585.12167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mekonen T, Ford S, Chan GCK, et al. What is the short-term remission rate for people with untreated depression? A systematic review and meta-analysis. J Affect Disord. 2022;296:17–25. doi: 10.1016/j.jad.2021.09.046. [DOI] [PubMed] [Google Scholar]
- 28.Vigod SN, Rochon PA. The impact of gender discrimination on a Woman’s Mental Health. EClinicalMedicine . 2020;20:100311. doi: 10.1016/j.eclinm.2020.100311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Patwardhan V, Gil GF, Arrieta A, et al. Differences across the lifespan between females and males in the top 20 causes of disease burden globally: a systematic analysis of the Global Burden of Disease Study 2021. Lancet Public Health. 2024;9:e282–94. doi: 10.1016/S2468-2667(24)00053-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Höhn A, Gampe J, Lindahl-Jacobsen R, et al. Do men avoid seeking medical advice? A register-based analysis of gender-specific changes in primary healthcare use after first hospitalisation at ages 60+ in Denmark. J Epidemiol Community Health. 2020;74:573–9. doi: 10.1136/jech-2019-213435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Harris MG, Baxter AJ, Reavley N, et al. Gender-related patterns and determinants of recent help-seeking for past-year affective, anxiety and substance use disorders: findings from a national epidemiological survey. Epidemiol Psychiatr Sci. 2016;25:548–61. doi: 10.1017/S2045796015000876. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data are available in a public, open access repository.