Malawi faces several challenges in implementing digital health programmes that facilitate health data recording and reporting. While the number of people with access to the internet in the country has increased from 9.6% in 2016 to 13.1% in 2018, overall access remains significantly low and Malawi still has very poor and substandard technological infrastructures ²³ . The country has the lowest internet access in comparison with regional neighbours such as Zambia where those with access to the internet are more than twice that of Malawi. According to the Inclusive Internet Index 2020 report, Malawi ranks very poor on all four of its indicators: internet availability, affordability, relevance, and readiness with the number of Mbps in Malawi regarded as the lowest and the least growing rate in the world. Moreover, high costs hinder access to necessary IT infrastructure and Internet for many users in developing countries like Malawi ²⁴ .

To date Malawi’s health system remains highly dependent on a paper-based approach, with paper registers and monthly reporting forms being the common data capture and reporting tools in many health facilities. Using paper-based reporting tools within health service delivery leads to many problems such as difficulty retrieving patient records, data inconsistencies, lacking adequate physical space to store paper records, costly maintenance of such records, lacking security and confidentiality of patient data, inability to access patient records from different points of care, poor integrity of patient information and lacking continuity of patient care ²⁵ . Malawi’s Ministry of Health started working on an eHealth strategy in 2014 that aimed to build foundations for ICT infrastructure, and implementation of eHealth solutions including individual electronic health records. Progress regarding electronic medical records (EMR) systems has been made in certain clinical areas like maternal and child health, HIV/AIDS, and TB. Sadly, EMR developments within mental health and related substance use services remain under prioritised on this front. Similarly, St John of God used the paper-based approach since its establishment in 1995. The digital ICT capturing of data started much later, with more prevalent use and investment in ICT taking shape around 2008.

Table 1 provides a summary of the resources and information systems available to capture epidemiological and health service delivery substance use disorders data in Malawi, based on 2014 records from the World Health Organisation (WHO) Global Health Observatory data repository. Broadly speaking the table illustrates severe lack of recorded information regarding substance use in the general population, lacking treatment programmes for adults and young people. According to available reported data from WHO regarding service delivery for substance use, treatment coverage for substance dependence is very limited, ranging between 11 to 20% for alcohol dependence and one to 10% for cannabis dependence. There is no data for cocaine dependence or opioid dependence in terms of treatment coverage. There is insufficient data to facilitate population level analyses of drug or alcohol as risk factors for mortality. These gaps in information further propagate lacking evidence regarding real prevalence of substance use issues in Malawi, and the true (economic) impact of substance use within an already overburdened and under resourced mental health service.

In this paper, we set out to analyse routinely collected data from outpatient and in-patient registry of two mental health services run by Saint John of God hospitaller Services (SJOG). These services are set in two geographical sites. The service in Mzuzu (northern region) was the first to be established in 1995 and the second is based in Lilongwe, Malawi's capital city, established in 2012 following increasing demands in mental health service provision ( Figure 1). We specifically report incidence of service users attending these two services for substance abuse treatment over a 9-year period (2010 to 2019).

Patient consent for this publication is not applicable as we used retrospective anonymised routinely collected data. Therefore the need to obtain informed consent from individual patients was not necessary and was waived by SJOG’s reviewing committee. Access to data used in this analysis was reviewed and approved by the St John of God Clinical Director on 20 ^th February 2020.

This analysis used routinely collected data related to substance use disorders from two SJOG hospital sites that provide psychosocial or psychopharmacological treatment emanating from primary assessment and diagnosis at the mental health clinic. Mental health and psychosocial support services, also known as psychosocial counselling where a professional facilitates actions that address both psychological and social needs of individuals, families and communities ²⁶ , is a relatively new phenomenon in Malawi. The average Malawian will rarely take voluntary action to go and pay someone to talk to them about their issues (talking therapies). A key reason for this is that talking to someone about something is anecdotally believed to be at the heart of Malawian culture and social life. Family, friends and church members still form a big part of the social support network for people to offload their issues, including mental health and substance use problems ²¹ . The recent movement towards professional counselling in Malawi is, in practice, an outgrowth of HIV testing and counselling ²⁷ . It was the gaps identified within HIV Testing clinics with regards to addressing mental health and psychosocial effects of HIV that inadvertently led to growth of the ‘psychosocial counselling’ field ²⁶ . However, more broadly counselling is still perceived as a western thing or something for the privileged. There is vital need to increase understanding about what therapy is beyond the common view that it is just talking to someone about your problems ²⁸ .

There are several routes through which one can access counselling services at SJOG centres in Lilongwe and Mzuzu:

(1) Complementary service, to an individual and family at the point of assessment for a mental health condition, admission (counselling is provided at individual and group levels in-patient care and in addiction recovery services) and subsequent post-discharge visits (as a way of exploring risk factors or protective factors prior to and after the mental illness with service users).

(2) Involuntary initiative, taken by family members for a person with poor insight to their problem (often related to personality traits and substance use); and/or by organisations for their employees.

(3) Provider-initiated process (PIP), mainly through the HIV testing and counselling service at SJOG. The testing centres were established to strengthen the mental health psychosocial support through seeing testing behaviour in a holistic manner by assessing multiple psychosocial related to HIV testing – i.e., exploration around underlying motives/drivers of multiple testing within shorter periods of time.

(4) Voluntary basis; sprung from a recent growing trend of voluntary counselling among the middle class and the younger more highly educated demographic of Malawi due to issues of debt, COVID-related mental health problems, increasing awareness of suicide cases, anxiety and depression, and the growing problem of substance abuse among their peers in the country ^{21, 29} .

All the services are provided at a cost.

Routinely collected anonymised patient data from the two SJOG hospital sites was descriptively analysed. Specifically, the number of service users per year presenting to the SJOG psychosocial services were tallied up based on the substance use disorder they were diagnosed with. These numbers started from service inception (2010) to the most recently available data (2019). It is worth noting that alcohol and tobacco use is confirmed by way of qualitative interview with the service user during clinic sessions and/or health record history without testing because they are legal substances. ‘Mixed-substance use’ disorders typically represents alcohol, cigarettes, and cannabis.

Figure 2 below shows the number of service users presenting to SJOG psychosocial services and being diagnosed with a substance use disorder per year from 2010 to 2019. The most common substances of use encountered at SJOG over the years both by way of history and testing are: (1) Alcohol (2) tobacco smoking (3) cannabis, locally called ‘chamba’ (4) pethidine (5) Others such as cocaine.

Over the 10-year period since service inception, SJOG saw 611 service users with varied substance use disorders for psychosocial treatment. Numbers of people with substance abuse presenting to the service generally increased from 2012 to 2019 with large increases particularly noted in 2014 (13-fold increase from 2013) and 2018 (3-fold increase from 2017). Alcohol withdrawal and chamba induced psychosis remained at the same level from 2010 to 2019. Prevalence rates of alcohol induced psychosis remained the same until 2018 when cases increased nearly 4-fold (15 to 60 service users) before reducing in 2019 but remained higher than previous years (34 service users).

Mixed substance abuse was a new category created in 2019 to capture the increased use of multiple substances (typically alcohol, chamba and tobacco) among service users undergoing treatment. This might explain some of the drop in alcohol and substance abuse case numbers in this year compared to 2018 as people have been recategorized to reflect the more complex issue of using multiple substances together rather than just concentrating on one main substance.

There are a few limitations for consideration when interpreting the findings in this present analysis. Data presented depicts retrospective aggregated records from paper files on attendance of service users to SJOG psychosocial services for substance use disorders treatment since service inception. We were therefore unable to detect presence of repeat users. However, a previous study investigating factors associated with readmission to inpatient psychiatric services in Lilongwe, Malawi found readmission was generally low (6.9%) and related to medication non-adherence ³² . Considering the relatable context of these two mental health services, we can assume similarly low repeat use or attendance of SJOG outpatients’ psychosocial services. In 2015, there were no records relating to diagnoses therefore we were unable to include this year within the analysis. This may underestimate the attendance rates to SJOG psychosocial services. Moreover, age was not routinely collected within the service up until 2018, therefore the numbers reported from 2010 to 2018 may include both adult and younger service users attending the service due to a substance use disorder diagnosis. As previously shown, information systems for collecting substance use data in Malawi are mostly lacking or very limited. This poor infrastructure limits our ability to make broader comparisons and/or more in-depth analyses of the SJOG data in relation to the whole country and beyond. Finally, data used for this analysis was from a specific organisation delivering mental health care for a specific health disorder. This organisation is privately run therefore may not include a representative sample, limiting generalisability of our findings.