Between Donor Interest, Global Models and Local Conditions: Treatment and Decision-Making in the Somalia-Finland Tuberculosis Control Project, 1981–3

Abstract

Despite numerous global health initiatives after World War II, tuberculosis still poses a major threat in sub-Saharan Africa. This article examines one attempt to tackle this problem: the Somalia-Finland Tuberculosis Control Project. Conducted in the 1980s as a bilateral development aid project between the two countries, it became the most extensive – and expensive – tuberculosis initiative in Somalia in that decade. An interesting feature of the project is that, despite a lack of previous experience in tuberculosis work in developing countries, the Finnish partner decided not to follow the WHO global guidelines designed to standardise tuberculosis activities across the developing world. Instead, Finns established their own treatment programme based on X-ray and short-course chemotherapy – otherwise rarely used in clinical practice in Africa. Through a close reading and comparison of the correspondence, project plans, memos and minutes, the article analyses the formation of this strategy. Focusing on ground-level decision-making, it argues that the decisions were based not only on a belief in the superior clinical effectiveness of these methods, but also on the fact that they better suited Finnish ambitions and project logic. Thus, the article supports the notion that donor perspectives on resources and project objectives determined what was seen as feasible treatment in a developing country. By shedding light on the debate between the supporters of short-course chemotherapy and the WHO standard treatment strategy, it also contributes to the early history of DOTS (directly observed treatment, short course).

In April and early May 1980, two Finnish tuberculosis specialists – Dr Eljas Brander and Dr Pähr-Einar Hellström – visited Somalia for two weeks as guests of the Somali Ministry of Health. As board members of the Finnish Anti-Tuberculosis Association (FATA), their mission was to collect information on the tuberculosis problem ¹ and investigate how FATA might support Somalia in tuberculosis control. ² The result was the founding of the Somalia-Finland Tuberculosis Control Project. ³ Conducted over the decade 1980–90 as the first bilateral aid programme between the governments of the two countries, the project displayed the key characteristics of a post-war global health initiative: it was a disease-specific intervention that was carried out in Africa by agencies based in Europe, and it was built upon a general belief that the disease burden in developing countries could and should be diminished by means of know-how and material resources from the West. ⁴ The Project received government-funding from the Finnish Department of International Development Cooperation (DIDC), and it remains one of the largest health aid initiatives that Finns have undertaken in developing countries. With a total budget of 43.5 million Euros (if converted into today’s currency) and over 35 000 cured patients, it was also the most extensive tuberculosis initiative in Somalia during the 1980s. ⁵

Finland did not have a past as a colonial power, whereas other European countries did and that was what motivated them to prolong their influence in Africa – also under the aegis of development aid. ⁶ Nonetheless, the project was still clearly tied to Finnish agendas. FATA was a major Finnish non-governmental public health organisation. It had been founded in 1907 for tuberculosis control in Finland, ⁷ where the disease (henceforth TB) had remained a major public health issue for much longer than in the other Nordic countries. ⁸ Once TB was no longer a threat at home, the Association turned its interests towards developing countries. Apart from the humanitarian global health goals, development aid allowed FATA to continue practical anti-TB work, upon which its self-image, influential position and funding was highly dependent. ⁹ The same move had already been seen in many other countries, for example Sweden, Norway, France and the UK. ¹⁰ The Somalia Project was FATA’s first development aid project. Previous contacts between Finland and Somalia had been modest and the Project began the bilateral co-operation between the two countries. ¹¹ The long-term objective was to assist the Somali National Tuberculosis Control Programme (NTCP) – under the authority of the Ministry of Health (MOH) – in its attempts to expand TB services and reduce the high prevalence of the disease in Somalia. In 1981, the immediate objective was to develop feasible case-finding and treatment measures in the Qoryoley refugee camps. ¹² From 1984 onwards, the Project evolved from a treatment-oriented initiative to an extensive healthcare infrastructure and education programme, operating in five of the sixteen regions of Somalia before it was suddenly terminated in 1990 with the outbreak of the Somali Civil War. ¹³

A historically interesting feature of the Project is that, despite having no previous practical experience of TB work in developing countries, FATA decided to deviate from the twelve-month standard treatment for tuberculosis recommended by the World Health Organization (WHO) – designed to be followed throughout the developing world, and already adopted earlier by the Somali NTCP. Instead, the Finns opted for a case-finding method that relied on X-ray scanning and a drug treatment of rifampicin and varying combinations of streptomycin, isoniazid, pyrazinamide and ethambutol over a period of six months, alongside a course of chemotherapy. During the early 1990s, the six-month chemotherapy would become one of the cornerstones of the ‘DOTS strategy’ (directly observed treatment, short course), adopted by the WHO in 1994 to be implemented globally. The strategy is based on five elements: (i) a political commitment to tuberculosis control; (ii) the diagnosis of cases based on sputum microscopy; (iii) standard treatment with a short course of directly observed chemotherapy during at least the initial phase of treatment; (iv) a regular drug supply to ensure the uninterrupted treatment of patients; (v) monitoring programmes using a standard recording and reporting system that includes reports on treatment outcomes. ¹⁴ This approach was developed by the International Union Against Tuberculosis (IUAT) during the 1980s; ¹⁵ however, as this article will demonstrate, the applicability of the short-term treatment in developing countries was still controversial in the early 1980s. The X-ray scanning was at odds with the case-finding principles of both the WHO and IUAT.

This article asks how the Finns chose and implemented this approach in Somalia in 1980–3. Explanations are sought by analysing the experiences, interests and opinions of Somali and Finnish participants; exploring the inner dynamics of the project; and looking at the international discussion concerning tuberculosis control. The article will first outline the context by mapping the background of Somalia’s NTCP, the tuberculosis situation there and how global tuberculosis plans impacted on Somalia in the early 1980s. Then it will examine the reasons why the Finns initially diverged from the standard WHO approach. Lastly, the article will show how short-term treatments were implemented in practice in Qoryoley, and how these experiences influenced treatment strategy in 1982–3.

The sources include both semi-official correspondence between the key Finnish actors and official documents such as plans, reports and meeting minutes from the private archive of Finland’s Lung Health Association, and governmental archives of the Finnish Ministry for Foreign Affairs. ¹⁶ The global tuberculosis guidelines are drawn from previous historical research and WHO documents available in the WHO IRIS database. ¹⁷ I will also refer to the interviews of five Project members, who worked in the field or in the FATA office in the period 1980–3. I interviewed them during the autumn of 2017. ¹⁸ My main focus, however, is on the letters written by the leading doctors in the Finnish medical team sent to Somalia, Dr Tor-Leif Sellergren (1981), Dr Eeva Rusanen (1982), Dr Liisa Parkkali (1983) and Dr Eljas Brander, the latter of whom would become the director of FATA in 1984. ¹⁹ Judging by this correspondence and interviews with former Project members, he was one of the main architects of the Project and acted as mediator between the medical team and the FATA board in 1981–3. ²⁰

The letters form a rich and immediate portrayal of the daily activities, hardships and cultural encounters made in the field. Like any sources, however, they also have their shortcomings. First, not every discussion about treatment left a written trace, since aid efforts were often planned over the phone and in unofficial meetings. ²¹ Second, the views of the Somalis are clearly underrepresented in the written sources. The historian Sanjoy Bhattacharya has pointed out that the reliance on sources produced by people in senior positions in international non-governmental organisations (NGOs) and national governments has caused medical historians to downplay the importance of local agency and ignore contradictory visions, and therefore diminish their impact on global health initiatives. ²² Although the article aims to map out the diversity of opinions and actions of the Somali side, this critique can be fairly levelled at this study, too. The Finnish archives used to contain only a few documents written by the representatives of the NTCP and MOH, and the ones I have found do not discuss treatment. In a few instances, their opinions have also been referred to in the letters of Finnish doctors. Thus, the article analyses the formation of the treatment strategy mainly from the donor’s perspective, moving between the field (in Somalia) and the central administration (in Finland).

The article makes two main contributions to the historiography of tuberculosis and global health interventions. First, by offering a detailed description on how short-term treatment was chosen, discussed and implemented at the ground level, it sheds light on the early introduction of short-course chemotherapies in Africa, which have been called ‘the last major breakthrough in TB control’. ²³ A recent historical interest in global, post-war tuberculosis interventions has focused on important developments in South Asia, mainly prior to the 1980s. ²⁴ Christoph Gradmann has analysed the IUAT-supported Tanzanian National Tuberculosis and Leprosy Programme, which served as the main inspiration for the DOTS that continues to be the cornerstone of global tuberculosis-control efforts. ²⁵ Otherwise, little is known about early programmes using short-course regimens in the African context. ²⁶ By offering a parallel model for short-course chemotherapy, the article also adds to the development of DOTS, which – despite good cure rates – has also been accused of being too top-down oriented and over-simplified for various conditions in the developing world. ²⁷

Second, this article adds to the historiography of post-war global health interventions by investigating disease control in bilateral development aid settings. As Bhattacharya has pointed out, WHO officials could not directly instruct donors and national authorities in how to implement bilateral health projects. ²⁸ Packard has noted that donors bypassed the WHO policies if they were not suitable for their own agendas. ²⁹ Donor-based planning and lack of co-ordination between different aid efforts have contributed to the fact that health care interventions have often remained vertical and treatment-oriented despite the attempts of the WHO to create more holistic, basic health care centred approaches to health. ³⁰ By examining how the ‘Finnish model’ of TB control was imported, how it intersected with the WHO policies and contemporary trends of TB control, and how the clinical work balanced between the high medical standards and realities of the local conditions, the article sheds more light on the logic of bilateral health care projects and their relations to the global guidelines.

1 Tuberculosis in Somalia

Although the project was initiated by FATA, the Somalis also had an obvious interest in collaborating. According to an estimate, one per cent of the general population and ten per cent of the refugees in Somalia were bacteriologically positive and thus infectious. Addressing TB was one of the main targets of the MOH’s five-year healthcare plan for the years 1981–5. ³¹ It coincided with the waning of international interest in anti-TB work, ³² having once been one of the main targets of the WHO and the United Nations International Children’s Emergency Fund (UNICEF) in the 1950s and 1960s. ³³ The first vertical anti-TB campaign had been implemented in Somalia in 1953 by the Italian administration of the UN trust territory of Somaliland. ³⁴ The WHO and UNICEF supported the establishment of treatment centres for TB in Hargeisa in the late 1950s, and in Mogadishu just after independence in the early 1960s. They also funded the establishment of the Somali NTCP in 1971 and its implementation in the years following. ³⁵ Such national programmes had become the WHO flagships for TB control in Africa. ³⁶ In Somalia, the aim was to expand TB services nationwide by establishing national tuberculosis centres and walk-in or ‘ambulatory’ treatment centres in each district. ³⁷ By 1980, however, TB had become a neglected disease due to the WHO running into financial difficulties, the disappearance of TB from developed countries, which decreased the general interest in the disease, disappointment with the results of TB control efforts and the Alma-Ata Declaration in 1978, which prioritised the development of primary healthcare (PHC) over specialised vertical programmes concentrating on one disease. ³⁸ The governmental agreement between Finland and Somalia notes that although there were many donors for public healthcare, there were none for tuberculosis services. The WHO and UNICEF were planning to discontinue their aid to the NTCP after 1980/1. ³⁹ Thus, a new donor was needed.

As a bilateral aid initiative, the Project was officially based on mutual negotiations between Finnish and Somali parties. ⁴⁰ The key Somali experts were the director general of MOH, Dr Mohammed Muse Gure, and the director of the NTCP, Professor Yusuf Hersi. They made it clear that Somalia wanted assistance with expanding NTCP among the Somali population. They requested the construction and renovation of TB centres (one in every region), training for anti-TB medical teams and material support in terms of medicine, equipment and transportation. ⁴¹ FATA, however, had planned to start the Project in the Qoryoley refugee camps. The camps had been established after the Ogaden War between Somalia and Ethiopia in 1978–9. The refugee situation had already received a lot of international attention, as the Somali government had recently requested international emergency aid for the camps. ⁴² There were officially 70 000 refugees in Qoryoley, ⁴³ but in 1981 Sellergren estimated the number to be only around 4500. ⁴⁴ Hellström and Brander drafted the first Project plan for refugee aid immediately after their visit and handed it to the Finnish Ministry of Foreign Affairs on 11 June 1980. ⁴⁵ This was revised by FATA in July. ⁴⁶ The governmental agreement between the Finns, Hersi and Muse Gure, plus the plan of operation, were drafted in August 1980. ⁴⁷ The long-term objective was to support the NTCP – demonstrating that both the Finns and the Somalis wanted to continue aid after 1981 ⁴⁸ – but in practice, FATA’s earlier plans from June and July were carried out. This prioritised the detection and treatment of tuberculosis in the Qoryoley refugee camps by the Finnish team (nurse, laboratory technician and X-ray technician led by a doctor), and to give practical training to their Somali counterparts. ⁴⁹

In the 1980s, the two most influential players in global TB policy were the WHO and the IUAT. The WHO Expert Committee’s periodic reports were seen as ‘guides to the global TB problem’. ⁵⁰ IUAT, an international NGO for tuberculosis control and umbrella organisation for national tuberculosis associations around the world, was one of the few organisations that still focused on TB in the 1980s, when most donors had lost interest in the disease. ⁵¹ Like the majority of African countries, ⁵² the national tuberculosis programme in Somalia followed the WHO standard treatment: patients were diagnosed using sputum microscopy and given a long-term combination course treatment of inexpensive drugs such as streptomycin, isoniazid and thioacetazone for twelve months. ⁵³ IUAT’s Tanzania programme also followed these principles in 1977–81. ⁵⁴ Based on the studies undertaken in Madras (1956–9) and Bangalore, India, the WHO standard treatment was intended as a cost-effective means for combating TB. ⁵⁵ The challenge with TB is that it is a disease closely tied to social and environmental conditions. Inadequate living and working conditions, overcrowding, malnutrition and poverty – typically related to developing countries – all increase the risk of infection and the chances of a latent disease becoming active. Matters are often made worse by the lack of healthcare infrastructure and material resources to treat the large number of patients falling ill under such conditions. ⁵⁶ The WHO TB control policy had prioritised the Bacille Calmette-Guérin (BCG) vaccination campaigns rather than treatment. ⁵⁷ Still, the results of the Madras study especially were regarded as ground-breaking because they indicated that patients living in poor socioeconomic conditions could be cured with cheaper ambulatory treatment and multi-drug chemotherapy instead of expensive hospitalisation, which was the norm in the West. ⁵⁸

However, it was not possible to reproduce the encouraging results of the closely supervised Madras trials in various real-life conditions elsewhere in the developing world. ⁵⁹ According to Sunil Amrith, the studies conducted in Bangalore already suggested that, although poor socioeconomic conditions did not affect clinical effectiveness per se, they did have a significant impact on the patients’ chances of completing their treatment. ⁶⁰ Brander’s travel report demonstrates that both Somali and Finnish tuberculosis specialists were aware that Somalia was a prime example of the typical problems that could befall WHO ambulatory treatment. It was demanding for both the patient and the provider, since it required regular visits to a clinic for twelve months. Patients often ‘defaulted’; in other words, they stopped the treatment too soon. The prevalence of nomadic culture in Somalia also made the treatment difficult to follow through, as did the lack of ambulatory treatment centres, drugs, transportation and staff. ⁶¹ Defaulting was a problem for two reasons: first, patients stayed infected and infectious; and, second, it encouraged drug-resistant strains of tuberculosis, thus rendering the standard drug regimen ineffective. ⁶² The fact that these TB drugs were also freely sold in pharmacies only exacerbated the problem further. ⁶³

Clinically-oriented TB specialists were already concerned about this in the early 1960s. ⁶⁴ The long duration of the treatment was seen as one of the main obstacles, and so the British Medical Research Council (MRC) carried out trials for short-course therapies in East Africa (Studies X and R), Singapore and Hong Kong during the 1970s. ⁶⁵ By the 1980s, nobody doubted that they were clinically more effective. In 1982, the report of a study group of the WHO and IUAT (1981) declared that all countries should have access to short-course therapies. Six- to nine-month regimens of rifampicin made patients non-infectious within three months, and this cut down the chain of infection quickly, and also proved effective against TB strains previously resistant to the twelve-month treatment. However, the organisations also stated that the WHO standard treatment should remain a basic regimen in many developing countries. ⁶⁶ One reason for this was the cost. The short-course regimens violated the PHC policy of the WHO for being too expensive for developing countries to be carried out independently. ⁶⁷ This was also the case in Somalia, as Hersi had pointed out to Brander and Hellström that the NTCP could not afford rifampicin. It was therefore prescribed only to those patients who could pay for it themselves. ⁶⁸ The WHO stipulated that the price of the regimens should come down significantly before they could be used. ⁶⁹ Another serious concern was that difficult clinical conditions might also lead to a rise in rifampicin-resistant strains, as new regimens had not yet been clinically tested in African countries. ⁷⁰ Both the WHO and IUAT also rejected using mass X-ray scanning for active case-finding as it was regarded too expensive, complicated and imprecise compared to sputum microscopy, thus leading to wrong diagnoses and significant over-reporting of the cases. ⁷¹ Instead, they recommended examining the sputum of symptomatic patients that independently came to the TB centres; and only bacteriologically positive infectious patients would then be treated. ⁷²

Despite the above, the WHO and IUAT clearly had different views on testing short-course chemotherapy in clinical practice. Due to the poor treatment results with the WHO standard treatment, IUAT started to use the eight-month regimen in clinical practice in Tanzania in 1982. The regimen was based on the research of the MRC from the 1970s. As Gradmann has pointed out, IUAT’s decision ran counter to the WHO recommendations caused a serious stir between the two organisations. ⁷³

2 Exporting the ‘Finnish model’

Whereas IUAT’s Tanzania programme was first based on the WHO standard treatment, Finns ignored the global strategy right from the beginning. According to Brander, there were two reasons for this: direct funding from the Finnish government, and the resistance situation. ⁷⁴ Brander and Hellström had initially planned to rely on the WHO recommendations, ⁷⁵ but during their visit they collected thirty-six sputum samples for sensitivity tests back in Finland. These took place in June 1980 and results indicated a high level of primary resistance to streptomycin and isoniazid. ⁷⁶ These were the primary drugs used in the WHO standard regimen, suggesting that, even if patients stayed in treatment for the full twelve months, many would not be cured. The duration of the treatment was also considered too long; and the second plan (made in July) noted that treatment could be reduced to eight or nine months and use more effective and expensive drugs. ⁷⁷ To avoid resistance, the drug combination had to include at least two drugs to which patients were found to be susceptible. By the autumn, FATA had decided to use a nine-month regimen starting with four drugs taken daily (rifampicin, streptomycin, isoniazid, pyrazinamide), ⁷⁸ with a follow-up phase where the drugs were taken only three times a week. ⁷⁹

The second plan was to base case-finding on mass screening with a-miniature X-ray rather than just sputum samples: only those cases revealing pulmonary shadows in their X-rays would then be verified bacteriologically by sputum microscopy. Both bacteriologically negative and positive cases would be treated. The plan was to keep tabs on the resistance situation by sending sputum samples to the Public Health Laboratory in Helsinki for sensitivity tests. ⁸⁰ Jones connects similar mass-screening preferences in Sri Lanka to a clinical approach where physicians were committed to detect and cure all TB cases. ⁸¹ Similarly, FATA stated that, instead of ‘constructing extensive systems, where gaining success takes time and is uncertain’, ⁸² it was aiming for fast and visible results by using methods that were the quickest and the most effective way to tackle TB. ⁸³ The above stance on ‘extensive systems’ can also be read as an opposition to the PHC policy of the WHO that prioritised the integration of TB programmes to the public health care services. Like IUAT’s Tanzania programme, the Project paid some lip service to the PHC. ⁸⁴ In practice, both were carried out as vertical, specialised projects. ⁸⁵ As a bilateral aid, the Project was not obligated to follow any international guidelines. Finns clearly had little trust in the capability of PHC in its ‘mostly non-existent’ state in Somalia. ⁸⁶ Thus, vertical TB work was seen as a better guarantee for good cure rates than PHC integration.

It is important to notice that the short-term treatment became possible only thanks to the DIDC’s financial support. Brander stated that the second reason for FATA to use rifampicin-based regimens was that the DIDC had turned out to be more generous with government funding than they had first expected. ⁸⁷ FATA could afford to put medical effectiveness above cost considerations. As Sellergren put it in January 1981:

The patients we find should receive high-quality treatment, no matter the cost. Somali staff should be fully taught how to supervise the treatment. After we terminate the project, the Somalis can then decide if they can afford to continue this. At least they have learned the right way. ⁸⁸

Defining this ‘right way’ was based on prevalent development ideology that presupposed that donor countries such as Finland had superior knowledge and technology that could be transferred to low-income countries such as Somalia, which would in turn trigger development. ⁸⁹ FATA’s leading role in the strategy formation was strengthened by the fact that the DIDC did not take an active role in the Project prior 1984. FATA members and DIDC officials justified the Project with the conviction that Finnish medical experts were well-equipped and better trained to tackle the problem in a developing country, especially with their recent experience of successful TB control in Finland. ⁹⁰ During the most active stage of the Finnish anti-tuberculosis campaign, FATA had become both wealthy and influential, ⁹¹ but its board members had little to no practical experience of development aid. Only Hellström had direct personal experience, working as the doctor in Tanzania in 1969–70. ⁹² Of the other board members, the most active in the Project were Brander, the head of a department in the Public Health Laboratory in Finland; Professor Eero Tala, head of the clinic of respiratory diseases at Turku University Hospital; ⁹³ Sakari Härö, a doctor and senior civil servant at the National Board of Health; ⁹⁴ and Professor Jorma Pätiälä, FATA director, having recently retired as head of the clinic for respiratory diseases at Helsinki University Hospital. ⁹⁵

It is not surprising with these backgrounds, that X-ray scanning and the nine-month regimen were adopted, as these were based on clinical practice in Finland. FATA was clearly aware of recent international research on the shorter chemotherapy treatments and their pertinence for developing countries, ⁹⁶ and this further encouraged the Finns to rely on shorter regimens. But although the four-drug combination and intermittence tallied with international research, ⁹⁷ the duration of the treatment tallied more with Finnish tuberculosis work, where the nine-month regimen had been tested and used since the 1970s. ⁹⁸ And using X-ray scanning in TB control was not altogether unheard of in Africa; according to the WHO, it was ‘still used in some countries’ there in the early 1980s. ⁹⁹ Brimnes and Jones note that X-ray scanning was also preferred by many local clinicians in Sri Lanka and India. ¹⁰⁰ Yet in Finland the method had a long tradition; between 1952 and 1982 all citizens over the age of fifteen were invited for a screening once every three years. ¹⁰¹ Representatives of FATA compared Somalia’s TB situation to Finland’s in the 1930s, and justified the use of X-ray and directly observed treatment by referring to the recent experiences and improved TB situation in Finland. ¹⁰² The fact that the mass screening was also coming to an end in Finland also probably encouraged FATA to continue the practice elsewhere. ¹⁰³ According to Brander, with its soon-to-be superfluous X-ray equipment and technicians, Finland was suited to export this expertise to developing countries. ¹⁰⁴

The existence of reciprocity donor and recipient motives was not uncommon in global tuberculosis work, nor in medical development aid in general. ¹⁰⁵ Indeed, the transfer of Finnish methods is understandable also in the light of FATA’s recent international re-orientation. FATA had been an IUAT member since 1927 and collaborated closely with it since tuberculosis had receded in Finland. It had first participated in global tuberculosis aid by contributing funds to the IUAT Mutual Assistance Programme. ¹⁰⁶ In addition to giving fast and effective treatment to Somali patients, the methods also offered an opportunity for FATA to participate in testing and improving new treatment measures for developing countries. Soon after the work in the camps had started in 1981, Brander wrote to Sellergren that the results of X-ray scanning should be presented at an international conference to ‘show how the case finding should be conducted in developing countries, for example in refugee camps’ since ‘the WHO experts understand nothing of it’. ¹⁰⁷ Scientific publications by the Somalia project would be scant. ¹⁰⁸ According to Maija Kajava, the former office secretary of FATA, scientific interests were not the main motive for the Project. ¹⁰⁹ Since the Project was more about practical work than scientific data collecting, its results were mostly presented in IUAT conferences and international meetings rather than more strictly scientific arenas. ¹¹⁰

Both Hersi and the American medical experts that advised the refugee camp authority, however, were critical of FATA’s plans. ¹¹¹ The Finnish medical team arrived in Somalia in December 1980. ¹¹² On 3 February 1981, Sellergren wrote to Brander that he had been told that:

1. (a) X-ray machines should not be used at all in the case finding;

2. (b) Refugees should not be treated with anything other than WHO guideline drugs (streptomycin, isoniazid, TB1);

3. (c) One camp should not receive better treatment than the other camps;

4. (d) Refugee camps should not receive better treatment [than the rest of the Somali population]. ¹¹³

Since source material on Somali motives regarding the treatment is scarce, it is difficult to say for sure, why Hersi nevertheless quickly agreed to follow FATA’s methods. ¹¹⁴ Sellergen’s reasoning and promises might have had an effect. He wrote an open letter to Hersi where he justified the use of X-ray. ¹¹⁵ As a part of the settlement, they also agreed that part of the rifampicin would be used for chronic cases in the Mogadishu TB Centre, where Hersi was stationed. ¹¹⁶ Most importantly, however, Hersi was probably aware that the likelihood of him being able to change the Finnish methods into those recommended by the WHO were slim. Maria Eriksson Baaz and others have noted that, despite the rhetoric of partnership, the interests of the donor party often dominate in development co-operation. As a sponsor, the donor can set up ‘the rules of the game’. ¹¹⁷ In our case, this much was demonstrated already at the planning stage, since the immediate objectives followed FATA’s plans. The fact that it was also a short four-month contract might have created extra pressure. ¹¹⁸ It was evident that DIDC was interested in continuing the Project after 1981, ¹¹⁹ but since FATA had no practical experience in implementing aid and since the Project received significant funding from the Finnish government, DIDC wanted to first test the Project. ¹²⁰ Hersi thought that the continuation of funding would depend on the results from Qoryoley. ¹²¹ As Finland was the only major donor and the continuation of aid was uncertain, it is plausible that Hersi did not want to compromise the Project by arguing too much over treatment. The refugee aid period thus functioned as a way of forging a new aid relationship that would eventually fulfil more general needs of NTCP.

3 Towards international models

After six weeks of preparations, negotiations and familiarisation with the local conditions in Somalia, the Finnish team and their Somali counterparts started to X-ray refugees on 21 January 1981. ¹²² The first patients were admitted for treatment on 7 February. ¹²³ The day-to-day activities in the camps were administered by Tor-Leif Sellergren, a specialist in lung diseases who, among holding other shorter medical posts, had worked since 1965 as a doctor in a TB ward in Raseborg, Finland. ¹²⁴ Finns and the NTCP staff did clinical work closely together. During the first months, X-ray technicians Raimo Leisilä and Mohammed Cabdi carried out X-ray scanning. Laboratory technician Aniisa Saalax prepared and examined sputum. Samples were also delivered in Mogadishu, where laboratory technician Heli Oramies examined them with the local staff. Sellergren diagnosed patients with Somali doctors Abdullah Shek Mohamed and Moxamed Giuni. Nurses Tuija-Liisa Peltola and Asha Jama Gashan administered the drug distribution. ¹²⁵ Finns instructed Somali counterparts in TB work, but also received valuable advice on how practical matters should be arranged in the camps. ¹²⁶ However, the practical decision-making on case-finding and treatment was mainly in the hands of Sellergren. Despite their early differences, he and Hersi shared a mutual respect and seem to have had a good working relationship. ¹²⁷ However, Hersi was stationed in Mogadishu and, as the director of the NTCP and a professor at the University of Mogadishu, he had also other responsibilities on his plate. ¹²⁸ FATA’s board and the working committee held meetings only once a month ¹²⁹ – insufficient to deal with all that could crop up on a daily basis in that time. Brander explained to Sellergren: ‘It is difficult to get the Board members together’, and ‘besides, they cannot really give any absolute rules on how to treat patients’. ¹³⁰

FATA’s short-course chemotherapy guidelines were similar to the treatment principles that would later become the DOTS strategy. The case-finding was different, but otherwise principles such as fully supervising drug intake and following each case up with a patient register were alike. ¹³¹ However, in 1981–3, the Project was not able to follow these principles through in the field. Perhaps the most pressing reason for FATA launching the Project in Qoryoley had been that it expected the camps to be a more stable, ‘realistic’ place to administer and monitor the treatment than the rest of Somalia, where nomadism was common. ¹³² In practice, however, refugees moved in, out and around the camps. ¹³³ When people heard how effective the Finnish treatment was, they started coming to the camps from Mogadishu and nearby villages. ¹³⁴ By the autumn of 1981, the Project’s treatment stations were overcrowded, and the team had to establish additional dispensaries for non-refugees in Qoryoley town and the village of Shalambood. ¹³⁵ Sellergren suspected that many patients discontinued their treatment as soon as they started to feel better, ¹³⁶ so, in an effort to prevent mobile patients from defaulting, they were sometimes given drugs for a week or so. ¹³⁷ Medication was thus not always fully supervised. There were also cases of rifampicin being stolen. ¹³⁸

FATA’s advice to verify cases bacteriologically was also often disregarded and treatments were sometimes started without a proper sputum sample. Clearly, the Finns felt pressure to start the treatments fast, as not only was the initial contract period short, but case-finding and getting diagnosed patients into treatment lagged during the first months. ¹³⁹ After three months in the field, Sellergren explained the situation by stating that ‘in the beginning, we might have been focused too one-sidedly on launching the treatments. We wanted to make a good impression and it was also probably the humane thing to do’. ¹⁴⁰ The team probably wanted concrete results before the next negotiations to be able to vindicate their methods. Also, the tuberculosis was more severe than anything they had seen in Finland. ¹⁴¹ In Somalia, TB was taboo and unsupervised self-medication was common, so people only came to the clinics when the disease was already further on and more difficult to treat. ¹⁴² In Finland, forty per cent of the patients who entered treatment were bacteriologically negative and had been diagnosed on the basis of X-ray or clinical symptoms. ¹⁴³ The Finns, who were used to starting treatment without bacteriological verification, probably regarded it as unnecessary and unethical to wait for such verification when faced with the grim reality in the camps. Sellergren also noted that these hectic conditions explained why both Finnish and Somali team members often failed to collect or annotate the sputum sample properly. ¹⁴⁴ Although the WHO advocated the sputum microscopy as a simple technology, Jones has demonstrated that similar practical problems in collecting and examining samples were also common in Sri Lanka. ¹⁴⁵

In order to stop patients defaulting, Sellergren modified FATA’s standard four-drug regimen. According to him, the combination of rifampicin, streptomycin, isoniazid and pyrazinamide was too toxic for the weaker, malnourished refugees, who would suffer from side effects and give up the treatment. ¹⁴⁶ To avoid this, he instigated an initial clinical examination, where the patient’s overall condition and previous treatment history were all considered. By September 1981, Sellergren had fifteen different regimens in place. The most common were a three-drug combination of rifampicin, streptomycin and isoniazid (250) used in new cases, and a combination of rifampicin, isoniazid and pyrazinamide (151) for those patients who had been treated for tuberculosis before and might be resistant to isoniazid and/or streptomycin. ¹⁴⁷ During the follow-up phase, the patient was also given the drugs daily rather than intermittently. Sellergren was again worried that intermittence would disrupt treatments as, according to him, the refugees, who were used to daily treatment, did not ‘understand anything about it [intermittence]’. ¹⁴⁸ The duration of the treatment also depended on the patient’s condition, monitored via X-ray and sputum samples. Usually it would last for nine months, but six- to eight-month treatments were also accepted. ¹⁴⁹ These solutions were based on clinical experience gathered in Finland and practical experience from working in the camps. According to Dr Liisa Parkkali, the three-drug combinations above were all common in Finland. ¹⁵⁰ By prioritising treatment over bacteriological diagnosis, and by optimising it for every patient, Sellergren clearly approached the Project from a clinical perspective where curing every TB case was the priority. His stance was not extraordinary, as similar conflict between clinical perspective and policy formation had taken place in the Madras trials. ¹⁵¹ Sellergren’s approach can be explained by his experience as a clinician in Finland and the treatment-oriented goals of the Project, the success of which were measured by the number of cases detected, treated and cured.

In 1981, 11 337 people were examined by mass miniature X-ray and 1075 patients were admitted for treatment. ¹⁵² After the first year, the Project seemed quite successful. According to FATA, prevalence of active tuberculosis dropped from 8.5 to 3.1 per cent among those examined within six months. Control X-rays and bacteriological tests also indicated that many difficult streptomycin- and isoniazid-resistant cases had improved with the treatment. ¹⁵³ In 1982, the Project was extended to Mogadishu, as the Somalis had planned. ¹⁵⁴ The governmental contract for the years 1982–4, based on a proposal drafted jointly by Hersi and Sellergren, claimed the Project’s results proved the effectiveness of the Finnish method of controlling TB in the refugee camps, especially the usefulness of mass scanning with miniature X-ray. ¹⁵⁵ The results were presented at both the IUAT African region meeting in 1981 and the IUAT Conference in Buenos Aires in 1982. ¹⁵⁶

Experiences from the camps, however, also indicated that there were at least some grounds for the earlier criticisms. It had become clear that only a small proportion of patients could be treated with rifampicin – even within the specific Project area – as it was so expensive. In other words, the project had little to no impact on the overall epidemiological situation in Somalia. ¹⁵⁷ This was a problem when it was evaluated from a public health perspective, i.e. whether the Finnish approach benefited the whole population rather than a small number of individual patients. ¹⁵⁸ In March 1982, Dr Ron Waldman, one of the refugee health advisors, again bluntly criticised FATA for using methods that Somalia could not afford to continue, let alone expand, once the Finns left. According to him, the Refugee Health Unit (RHU) would never have accepted the Project’s treatment strategy. ¹⁵⁹ Dr Eeva Rusanen, who started her work as the team leader in 1982, was told by Brander to not be too ‘bothered about’ the above comments. ¹⁶⁰ Both FATA and Sellergren replied there was no point using the WHO standard regimen in Somalia as it was ineffective against resistant strains and the patients defaulted because the treatment was too long. ¹⁶¹ The experience from the camps had nevertheless shown that the Finns would need to modify their methods.

In 1982, the Finnish medical team started to work in Mogadishu, where the number of patients was expected to be even greater. ¹⁶² In early 1982, Hersi suggested that rifampicin be used only for chronic (resistant) cases, so that more patients could be treated. ¹⁶³ But, whereas public-health oriented experts considered it to be a problem that only small numbers were being treated, clinically-oriented FATA wanted rifampicin-based regimens as a standard treatment for every patient. It set new restrictions on the admission for treatment instead. From as early as March 1981, active case-finding was disregarded, and the treatment of bacteriologically positive cases prioritised – in line with the standard WHO treatment and the IUAT Project in Tanzania. ¹⁶⁴ In September 1982, Dr Rusanen wrote that the Finnish-Somali team at the UNICEF TB Centre in Mogadishu was limiting X-ray scans to fifty patients per day, and a Somali doctor would then choose two ‘suitable, preferably young and fresh cases’ for the Project. ¹⁶⁵ The rest were sent to NTCP for the WHO standard treatment. ¹⁶⁶ This method of case-finding thus moved towards more international models, as the importance of sputum microscopy had increased even if X-ray remained the scanning tool for sample collection. According to Liisa Parkkali, who became team leader in 1983, ‘clear cases’ were still admitted on the basis of X-ray diagnosis. ¹⁶⁷ Such restrictions were sometimes ethically difficult for the medical team to put into practice, ¹⁶⁸ but they made sense from the Project’s perspective: it had limited resources, but would be considered successful if the predetermined amount of patients were cured per year.

FATA was more concerned about defaulting patients than the number of treatments, as it directly affected treatment results, was a waste of expensive drugs and might eventually create rifampicin-resistant strains. According to Sellergren’s report from 1982, the defaulting rate had been at fifteen per cent in the Qoryoley camps in 1981. ¹⁶⁹ If this figure is accurate, the Project was extremely successful in reducing the rate of defaulting. In comparison, Dr Paul Shears calculated that fifty per cent of the 600 patients given the WHO standard treatment between July 1980 and August 1981 in the Somali Sab’ad refugee camps had defaulted, and of 244 patients still being treated, seventy participated only irregularly. ¹⁷⁰ It is unclear what data the Qoryoley figures are based on though; so far, I have not found clear statistical information on defaulters for the period 1981–3. Another report stated that, by September 1981, 120 of the total of 900 treatments started by that date had ended with defaulting and 170 had been successfully completed. ¹⁷¹ If we analyse only the treatments that had come to an end (170 cured + 120 defaulted  = 290), it seems that forty-four per cent of the patients had defaulted and sixty-six per cent completed the regimen successfully by autumn 1981. ¹⁷² According to Liisa Parkkali, the Finns (herself included) did not realise just how many patients were defaulting until they started using IUAT patient records in 1983–4. ¹⁷³

In August 1981, FATA decided to send the nurse Päivi Kokkonen to investigate the defaulting problem in the camps. ¹⁷⁴ Taking a public-health oriented stance, she recommended that, instead of trying to export ‘Finnish tuberculosis polyclinics’ to Somalia, the Project should improve treatment compliance by intensifying health education and community involvement at all levels, co-operating with primary healthcare services and using more Somali staff. ¹⁷⁵ FATA’s solution would remain mainly biomedical: from 1982, the Finns started using an even shorter, six-month course of chemotherapy in Mogadishu, the Qoryoley refugee camps and (from 1983) Kismayo. ¹⁷⁶ The new drug combination was based on international research (justified by Brander with the publication of Wallace Fox (1981) ¹⁷⁷ ). Drug-sensitive patients (new cases) received rifampicin, streptomycin, isoniazid and pyrazinamide daily for two months followed by isoniazid and rifampicin daily for four months. For drug-resistant patients (old cases) streptomycin was replaced with ethambutol. ¹⁷⁸ Concern about rifampicin resistance might have influenced the decision to return to a four-drug combination: streptomycin, rifampicin and isoniazid were a feasible combination in Finland, where resistant TB strains were uncommon, ¹⁷⁹ but in Somalia the same regimen could cause rifampicin resistance in those patients who were already resistant to streptomycin and isoniazid. In 1982, Brander further instructed that patients’ sensitivity to drugs should be ascertained before treatment was started and constantly monitored throughout. ¹⁸⁰

In 1981–3, the Finns focused on preventing the defaulting with flexible drug distribution. Despite the shorter treatment time, default rates were calculated to be standing at around thirty per cent in 1982 and 1983. Both Rusanen and Parkkali believed the long difficult journey to the clinics were the main cause, so it was common to give patients enough drugs for one to two weeks at a time. ¹⁸¹ An alternative solution would have been to restrict treatment only to people who lived near the ambulatories and could visit daily, ¹⁸² but this was rejected for both practical and ethical reasons. First, it was impracticable in a country that lacked medical services and, with no IDs, patient registers or even system of addresses, it would have been impossible to verify where the patient lived. Second, the idea of denying care based solely on where the patient lived did not tally with the Finns’ humanitarian conception of the Project. ¹⁸³ Local realities also caused other challenges. Treatments were disturbed when problems were encountered in regulating, ordering and transporting the drugs. ¹⁸⁴ Finally, the waiting time required for the sensitivity tests was so long that the clinic risked losing contact with patients, so treatments were started before results had come back and only changed later if there was drug resistance. ¹⁸⁵

The Project seems to be among the first to use the short, six-month course in clinical practice within a national tuberculosis control programme in Africa. While the new regimen was launched in Mogadishu in 1982, IUAT started its ground-breaking short-course chemotherapy pilot programme in Tanzania. The principles of the DOTS strategy were developed in that pilot programme under the leadership of the IUAT scientific director, Karel Styblo. ¹⁸⁶ The programme was extended district by district to the whole country in 1982–6. Collaborative IUAT projects were also started in Mali, Benin, Kenya, Malawi, Mozambique, Nicaragua, Senegal and Yemen. ¹⁸⁷ Although the FATA papers do not mention them, the Finnish TB specialists were probably aware of these developments. The IUAT leadership was certainly interested, and aware of the treatment carried out in the Somali-Finnish project. Styblo and Annik Rouillon, executive director of the IUAT, offered advice to FATA in late 1982, and FATA members enthusiastically took part in IUAT conferences and meetings. ¹⁸⁸ Procuring cheaper drugs through IUAT networks allowed FATA to increase the number of patients in the Project from 1000 in 1982 to 3000 in 1983. ¹⁸⁹

Although FATA was willing to collaborate with IUAT on its own terms, it wanted to stay in charge of the Project. This became evident when the DIDC asked Karel Styblo to evaluate the Project at the end of 1983. The small number of patients, the cost and the rifampicin resistance became pressing issues again. Both the Somalia-Finland Project and IUAT’s Tanzania programme were carried out as a specialised, vertical programmes based on high medical standards. However, their main difference was that the Tanzania programme had a country-wide approach on TB control. Thus, IUAT – like the WHO – prioritised the treatment of the masses. The Tanzania programme used the cheapest possible short-course treatment that lasted eight months and contained rifampicin only in the two-month long initial phase. ¹⁹⁰ The Somalia-Finland Project focused on curing a specific number of patients in the project area, and used the shorter and more expensive six-month regimen that contained rifampicin in both the initial and the follow-up phases. ¹⁹¹ Another significant difference was that the Tanzania programme prioritised the strict supervision of drug intake. Therefore, patients were hospitalised during the initial phase of IUAT treatment to ensure that taking the rifampicin was fully supervised. The hospitalisation was against the WHO guidelines that recommended ambulatory treatment for the developing world. ¹⁹² In the evaluation report, Styblo pointed out that the supervision had not been possible with FATA’s ambulatory treatment. He recommended that the Finns try out the cheaper IUAT regimen in routine conditions as a ‘pilot study’, but otherwise prioritise supporting the NTCP in improving their use of the WHO standard treatment, so that more patients could be treated. He reasoned that ‘it would take several years to know whether short-course chemotherapy is feasible in Somalia’. ¹⁹³ It is good to note that even the Tanzania programme could not carry out all its principles in clinical practice. ¹⁹⁴ Thus, the Somalia-Finland Project was not extraordinary in its incapability to fully follow high medical standards.

Styblo’s recommendations might seem odd in the light of his work in Tanzania, which challenged the WHO standard treatment. However, his stance might be more understandable in the light of his concern about the rifampicin resistance and costs. It is important to note that the short-course chemotherapy had been tested in Tanzania only approximately a little over a year when Styblo evaluated the Somali-Finland Project. ¹⁹⁵ Gradmann has stated that the Tanzania programme became particularly valuable because it had used both the WHO standard treatment and the short-course chemotherapy, and could therefore convincingly argue for the latter. ¹⁹⁶ Styblo clearly wanted this testing to happen in Somalia also before changing the regimen. He pointed out that there were no reliable, systematically gathered data on the results of the WHO regimen. He worried about the development of rifampicin resistance and was not fully convinced that Finnish-Somali teams could provide strictly supervised treatment in Somalia. He also evaluated the Project from the viewpoint of building a nationwide TB strategy. Since the Finns could afford to treat only a small proportion of patients with short-course chemotherapy, it made sense from the public health perspective to put more emphasis on the majority of the cases under standard treatment. ¹⁹⁷

FATA did not accept Styblo’s recommendation to focus on the WHO standard treatment. However, the Project started to hospitalise those patients who were at risk of default (i.e. lived far away from the TB services), supervised rifampicin intake more strictly and included eight-month treatment to the programme. These adjustments towards the IUAT model were highly influenced by the fact that they were promoted by DIDC, which took a more active role in supervision after the evaluation. Hospitalisation and supervision also made sense for the Finnish TB specialists, who saw defaulting as a problem. However, hospitalisation only became possible after 1983, when the DIDC decided to fund the construction and renovation of TB centres. The six-month course was still preferred by FATA and frequently used in daily supervision. Although most of the patients were sputum positive, sputum negative cases were also treated. ¹⁹⁸

It is important to notice that, after early 1982, there are no sources indicating that Hersi (or MOH) opposed the use of short-course chemotherapy. In the summer of 1982, Hersi actually defended the use of rifampicin-based regimens in the refugee camps, in a letter to the Finnish refugee council. In 1983, Dr Parkkali stated that there were no disagreements between herself and Hersi over medical issues. ¹⁹⁹ Since the Somalis involved in the Project did not shy away from expressing their differing views on other occasions, it is likely that they would have voiced their objections if they had disagreed about the treatments after the evaluation. It seems that the Somali leadership had developed a more positive understanding of the rifampicin-based regimens over time. On the other hand, the support may also have been first partly tactical, as a way of securing further funding for TB control. As the disease was not a priority in global health during the 1980s, it might have been difficult to find another donor.

4 Conclusion

The discussion on feasible treatment in the Somalia-Finland Tuberculosis Control Project demonstrates that the methods thriving in closely supervised trials and in industrialised countries were difficult to transfer successfully to the clinical practice in the developing world. The Project took place at a time when the short-course regimens were just beginning to be clinically tested. Finns participated in this work by exporting the Finnish model of TB control to Somalia. However, it soon became evident that this approach was not feasible as such in the local context. The Project could not fully follow high medical standards planned in Finland with limited practical knowledge of the local conditions. It had to adapt towards global models, especially towards the IUAT Tanzania programme, that were better suited for managing the oversupply of patients. However, despite criticism from the medical experts advocating the WHO or the IUAT model, Finnish preferences on effective but expensive practices such as X-ray scanning, treating bacteriologically negative patients and using the six-month regimen also remained an integral part of the Project.

The article shows that donor perspective on resources, objectives and local realities determined what was seen as feasible treatment. First, it argues that FATA’s decision to deviate from the WHO and IUAT models should be seen in the context of the broader international debate between clinical scientists and public health-oriented tuberculosis experts and organisations. The former, especially experts working directly in the field, put emphasis on cutting the chain of infection fast by curing individual patients. From this perspective, using regimens that were difficult for the patients to follow and caused resistance only made the TB problem worse. FATA’s decision had strong clinical grounds. Finns could afford to treat significantly fewer people with short-course therapy, but could also be relatively sure that the patients who stayed in treatment were cured. X-ray scanning and the shorter treatment regimen were also rational choices in the light of FATA’s past experiences in Finland and its interest in international TB work. FATA was a scientifically-oriented organisation dominated by clinicians. By implementing ‘new’ strategies, the Association could participate in the development of TB control models for developing countries and seek international recognition for ‘Finnish expertise’. Meanwhile, the public health approach, endorsed by the WHO, reviewed treatment on a mass level and aimed for nationwide coverage of healthcare services. Hersi’s approach was public health-oriented: as director of the NTCP, his main concern was to expand the treatment nationwide. X-ray screenings and an extensive use of expensive rifampicin did not seem like the best way to achieve this when only a fraction of the infected population could be treated.

Second, the article shows that the Finnish clinical approach was possible because the work was conducted as a bilateral development aid. Although the WHO TB strategy and PHC policy were understood as the official guidelines for the developing world, the Project could ignore them due to the extensive governmental funding, project approach and firm belief in Finnish medical knowledge. This gave the Project a strong identity as a vertical initiative operating under special agreement outside national policy lines. The project logic explains why FATA did not consider the high costs to be a major problem. Unlike the IUAT Tanzania programme, the Somalia-Finland Project was in practice a geographically and temporary limited intervention that focused on immediate, treatment-oriented objectives. The Finns nevertheless understood very well that their work covered only a small percentage of the cases and had, at best, only a minimal effect on the prevalence of TB in Somalia; but it was irrelevant to FATA at this point whether their methods could be extended fast to the whole country. The initiative was successful if the medical team could treat and cure a specific number of the patients within each contract period. From this viewpoint, the Finns were not responsible for the developments that took place outside the geographical or temporal scope of their project. The use of less effective methods made no sense to FATA or the medical staff involved with patients on a daily basis.