‘First do no Harm’: Deploying Professional Volunteers as Knowledge Intermediaries

Chapter 2 discusses the first part of our journey in operationalising the Sustainable Volunteering Project. It discusses the factors underlying the perceived ‘human resource crisis’ that is typically blamed for high levels of maternal and newborn mortality in low-resource settings. This is the environment within which professional volunteers find themselves and that they, and their deploying organisations, must negotiate with care. The chapter presents the risks associated with labour substitution or gap-filling roles and explains the importance of the co-presence principle to the SVP.


INTRODUCTION
Chapter 2 outlines the human resource context within which projects such as the SVP are deploying UK clinical volunteers. It begins with a brief presentation of global health 'metrics' emphasising the public view of the human resource crisis in LMICs. These stark metrics play an important (and intentional) role in stimulating the case for AID in all its forms including professional volunteering. Aggregate data on human resources in health form an important component of needs assessment. However, they are profoundly inaccurate in terms of conveying a statistical impression of health worker deployment on the ground due to the very poor and those with the greatest burden of disease must make do with a much smaller health workforce. The African Region suffers more than 24 % of the global burden of disease but has access to only 3 % of health workers and less than 1 % of the world's financial resources. 2 The clamour for metrics in the development/global health industry encourages the collection and aggregation of data which, perhaps unintentionally, drives policy agenda and intervention strategies. Table 2.1 summarises data from the WHO's 'World Health Statistics Report ' (2010).
It is important that we do not accept these figures as facts but approximations; numerous data bases report quite significant differences. However, the underlying message is clear: LMICs have far fewer skilled professionals than HRCs. In 2006, the WHO's World Health Report identifies a crucial threshold of 228 skilled health professionals per 100,000 population, below which countries were deemed to be in health workforce crisis (WHO 2006: 13).
Key stakeholders respond to this kind of data when designing their interventions. The Lancet Commission on Global Surgery 2030 (Meara et al. 2015) is just one example. Once again focused on 'global metrics', the Lancet Commission identifies five 'key messages', which include '5 billion people do not have access to safe, affordable surgical and anaesthetic care when needed' and '143 million additional surgical procedures are need in LMICs each year to save lives and prevent disability' (p. 569). On the basis of this, they identify six 'core indicators', the second of which is focused on improving workforce density: Kinfu et al. argue that the overall problem is 'so serious that in many instances there is simply not enough human capacity even to absorb, deploy and efficiently use the substantial funds that are considered necessary to improve health in these countries ' (2009: 225). Although they don't single out development aid, this statement may well apply to this form of funding too. Their analysis suggests that current figures may represent a marked underestimation of staff shortages. However, data weaknesses preclude accurate analysis and even regional data 'mask diverse patterns' (p. 226).
The data presented above and typically cited focus on 'stocks' (overall numbers) but tell us little about how the existing workforce is deployed and managed on the ground and how foreign human resource investments (in the form of foreign expertise) can best be managed.

THE HUMAN RESOURCE CRISIS IN UGANDA: CONTEXTUALISED KNOWLEDGE
The Ugandan Ministry of Health's Health Sector Strategic Plan III (MOH 2010) asserts that 'Uganda, like many developing countries, is experiencing a serious human resource crisis' (p. 20) restricting the country's ability to respond to its health needs. 3 It goes on to state that around 40 % of its human resource in health is working for the private sector (which includes the mission sector). One of the consequences of these shortages is a high proportion of unfilled vacancies in the public health sector. In 2008, only 51 % of approved positions were filled with vacancies reaching highest levels (67 %) in lower-level community-based facilities (p. 20). Facilities in urban areas and especially the capital city (Kampala) are less likely to experience problems with unfilled vacancies in comparison to more peripheral locations. The Strategic Plan reflects on the reasons behind this situation. And familiar concerns are raised over international migration ('brain drain') as health workers are attracted not only to resource-rich economies but also to neighbouring African countries such as Rwanda and Kenya where salaries are much higher and visas easier to obtain. Other factors identified include insufficient training capacity, low levels of remuneration (forcing forms of 'internal brain drain' or deskilling as qualified workers move to other sectors) and poor working conditions. However, even taking these factors into account does not explain the levels of staffing observed and experienced on the ground in Ugandan health facilities resulting in the pressures put on professional volunteers to gap-fill. The Strategic Plan goes on to identify low productivity as a result of 'high rates of absenteeism and rampant dualism' as the 'largest waste factor in the public health sector in the country' (p. 21). The World Health Report (WHO 2006) backs this up suggesting medical personnel absenteeism rates from 23 % to 40 % in Uganda (p. 190) and a World Bank Report (2009) quantifies the costs associated with absenteeism at UGX 26 billion. It goes on to identify the second most important source of waste as that arising from 'distortions from the management of development assistance', which constitute a 'major source of funding but are mainly off-budget' (World Bank 2009: 24). 4 The ubiquitous 'human resource crisis' is repeatedly referred to in research papers in the field of 'human resources for health' (HRH) but remains underspecified with vague references to an overall lack of personnel and/or lack of necessary training and skills (Thorsen et al. 2012). Indeed, it is hard to find a paper that does not refer to the lack of skilled personnel in facilities as a major factor. However, the reader is often left wondering what lies behind this situation and what it means in practical terms for health workers and, in our case, professional volunteers. Generic reference to 'staff shortages' tells us very little about the situation on the ground.
When asked to explain the reasons for staff shortages in Ugandan health facilities, an experienced Ugandan health professional replied: To start with really they don't have enough people trained to fill all the possible positions. I know that almost all the big hospitals are advertising positions for doctors and nurses. I also know lots of doctors who don't want to practice as doctors because they can work as consultants in an NGO. They usually go to American funders, they basically look around everywhere for anyone interested in funding their opportunities. People are now trying to go for project jobs. One good thing that people have realised is you can work in a government institution because there you are guaranteed a lifetime job and, at the same time, there are so many projects that come into the government institutions and help people top up their salaries in one way or another (UHW).
The respondent identifies a number of contributory factors. In the first instance, he indicates problems in initial supply exacerbated by the haemorrhaging of doctors from clinical work into (usually non-clinical) positions in NGOs. Others strategically seek to combine 'project' work with their full-time public roles (contributing to absenteeism and exhaustion). The respondent later refers to the problems of international brain drain suggesting that many Ugandan doctors are looking for better-paid work across the border in Rwanda, for example. But this is compounded by the often more damaging but neglected effects of 'internal brain drain' (Ackers and Gill 2008). In Uganda, this manifests itself in many doctors studying for Masters Degrees in either Business Administration (MBA) or Public Health (MPH), positioning themselves to work in NGOs in managerial positions. 5 Linked to the above, remuneration is a key factor affecting the presence of doctors in public health facilities. At the present time, private work ('moonlighting') is, in theory, illegal. In practice, it is endemic. To some extent, this represents a natural and entirely rational response to low pay. The following Ugandan health worker explains both the need for salary augmentation and the importance of holding a position in the public sector to facilitate private work: Most doctors working in the private sector are working for themselves simply because they need to make a bit of extra money and that way they can even negotiate to take some of the patients from the public hospital to their private hospitals (UHW) In reality, it is not so much that the private work 'tops-up' or brings in a bit extrathe balance is rather the other way around with private earnings dwarfing public sector pay. One specialist heavily involved in very lucrative fertility treatment referred to his public role as his 'charity work'. In other cases, doctors, most of whom do not own their own premises, clamour around NGO projects involving infrastructural investments in the hopes that the more attractive and functional facilities will enable them to attract fee-paying patients.
In addition to the low level of pay, serious administrative problems in many districts means that healthcare staff are not paid at all for months: Right now they are not paying them enough and it doesn't come on time. I know people who don't get paid for six months and they expect them to carry on smiling, offering the best services they can when their landlords are chucking them out because they don't have money to pay (UHW). This respondent had personal experience having waited for over 6 months to be paid (in this case by a university). Remuneration remains a major problem but it is never the only factor is not at all clear that a recent MOH initiative to significantly increase the pay of doctors in HCIVs (to 2.4 million per montharound £500) has translated into (any) increased presence on the ground.
In a rare study focused specifically on the absenteeism of health workers, Garcia-Prado and Chawla (2006: 92) cite WHO statistics indicating absenteeism rates of 35 % in Uganda. The reality is far worse. A senior manager of a Ugandan Health District reported (in an interview in 2015) much higher genuine rates of absenteeism, suggesting that during a personal visit that week, he found that over 65 % of his staff are 'on "offs"' at any point in time. This certainly confirms our experiences as ethnographic researchers and is likely to significantly over-estimate the presence of doctors. On one of the facilities we are currently involved with, the in-charge doctor has not been present at work for over 4 months (for no apparent reason).
Whilst overall health worker-patient ratios are relatively very low and many positions for which funding has been committed lie unfilled, it remains absolutely clear from our interviews and ethnographic work that the staff who are appointed and receive remuneration are very often not present for work. And the more senior the position the less likely they are to be present. In the following focus group with Ugandan midwives and doctors, respondents were asked about health worker absenteeism. They talked at length about midwives and nurses but did not mention doctors:

Interviewer:
You haven't mentioned doctors at all? (Laughter between everyone) Respondent 1 (midwife): Oh, sometimes we forget about them because most of the time we are on our own. You can take a week without seeing a doctor so we end up not counting them among our staff. Respondent 2 (doctor): Especially on a night, you never see them there (at the health centre).

Respondent 1:
Even during the day like most of the time.

Interviewer:
How often would you say a doctor would come to the facility in a typical month? Respondent 1: The medical officers have the rest of this centre to cover too so maternity will see them only if there is any problem. So they come for two hours three times a week but that's for the whole centre, the other wards as well.

Respondent 2:
Yes, like two times a week, sometimes once but most of that time even when they're on [duty] someone will not come to review the mothers.

Interviewer:
What would happen if a mother needs a caesarean? Would you call the doctor? Respondent 1: Initially they told us we should call before [referring] but every time you call that doctor he is going to tell the same thing: 'I'm not around, you refer'. And you use your own judgement but sometimes you follow protocol, because if anything happens . . . you call that doctor for the sake of calling.

Interviewer:
Just going through the process? Respondent 2: But you know he's not going to come (FG) In another location, the facility manager (a nurse) explains that, at the time of interview, there were few other factors restricting the use of the operating theatre (for caesarean sections): Now we have constant powerthe power is there. We had issues of water now they've stabilised. Now water is flowing; the issue of drugs, we have sourced drugs.
Interviewer: But the doctors are still not here? No, they don't even come and you have to keep calling. You will call the whole day and some will even leave their phone off. [ It is not simply that doctors work very few hours, but the unpredictability of their presence and the absolute resistance to commit to any set hours seriously impacts services and volunteer engagement. This situation has made it impossible for any of the facilities that we work in to run an electives caesarean section list, with the result that all cases become emergencies and are referred. 6 This not only causes serious delays for mothers but also makes it very difficult for professional volunteers to engage effectively with local staff and share skills for systems improvement.
Accommodation is a serious issue (as noted earlier), but it is not a panacea especially when it comes to doctors. In one case where our charity has funded a doctor's overnight room, it has yet to be utilised. On the other hand, where we have provided an overnight room for midwives (in another facility) we have achieved and sustained 24/7 working. Furthermore, in one of the health centres we are involved with where doctors benefit from the provision of dedicated (family) housing on site, this has not improved their presence. The following quote is taken from an SVP volunteer report: Caesarean section mothers operated on Thursday or Friday are generally not reviewed by a doctor over the weekend. One mother operated on for obstructed labour whose baby died during delivery had a serious wound infection, pyrexia and tachycardia and pleaded (4 days later) for me to help her (V). 7 Another volunteer made the comments in a report she drafted for the District Health Officer just before she left: Medical attendance or lack of it caused many problems. [ . . . ] in my own experience employed staff negate their responsibility when other professionals are on the ground believing that they will do their work and that they are free to work elsewhere (V).
She was referring here both to (foreign) volunteer presence but also to a visit by doctors from the National Referral Hospital during which time local doctors disappeared.
Whilst absenteeism and poor time-keeping are endemic problems amongst all cadres in Uganda, the situation is most acute when it comes to doctors. 'In-charge' doctors (senior medical officers appointed as facility managers) are often the worst offenders setting a very poor example to medical officers in their facilities and failing to observe and enforce contractual terms. As the following medical officer suggests, many if not most of the doctors in these leadership positions do not do any clinical work in the public facility they preside over: Most of the (in-charge doctors), if you really look at them, want to do administrative work actually, they want to sit in the officethey sign out the PHC (primary health care) fund. It's at their discretion to spend it so. . . . And of course sometimes there's corruption, outright corruption.
Interviewer: So really what they're doing is administration but not leadership.
Leadership requires you to be around; you can't let people run the place when you're not there. Leadership needs your presence, so you know the fact that [the in-charge doctors] are not always there, it's difficult. (UHW) Where in-charges are nurses, midwives or administrators, they have very limited ability to hold doctors to rotas: [Enforcement] is a problem. Doctors don't want to be accountable to someone 'below' them. They don't want someone, even if someone has a degree but they're not a doctor, to keep instructing them. (UHW) This problem of enforcement seems to stem from higher levels with District Health Officers (usually doctors themselves) seemingly powerless, or unwilling, to challenge poor behaviour: I think particularly in the health department they are still intimidated by doctors which is a bit surprising. It goes hand in hand with accountability because if I know I am accountable for something going missing and if it goes missing then something will be done to me; in terms of discipline then of course I will behave differently. I wouldn't want to be found doing something on the wrong side of the law because I know that there is action that is going to be taken against me. But because here people don't see anything being done then they can do lots of things. (UHW) A recent audit conducted by a volunteer of referrals to the National Referral Hospital from a Health Centre IV facility clearly identifies the problem of physician presence. It is important to point out that there are five physicians employed to work in this facilityfar more than most comparable health centres: Figure 2.1 shows that 62 % of referrals relate to human resource issues with 59 % directly attributable to the failure of doctors employed in the facility to be present during their rota hours. The situation reported here is by no means unusual. In one of the Regional Referral Hospitals we are involved with the professional (obstetrician) volunteer has instituted a weekly maternal mortality review process. On average two women die every week in this facility. The weekly reports highlight the human resource factors contributing to deaths. In most cases, medical interns are having to take responsibility for the bulk of referred patients despite the fact that the hospital employs four consultants. These consultants are rarely present It is also interesting to note that since the review process commenced, none of the consultants has attended the maternal mortality review meetings. It goes without saying that this situation has a very serious impact on health systems, intern supervision and patient outcomes. Its impact on the effectiveness of professional volunteer deployment is less well recognised. On the one hand, in an environment where absenteeism is neither recognised nor punished, the presence of skilled volunteers actually facilitates it. It is more difficult from an ethical and visibility point of view for a Ugandan health worker to leave a ward with no staff (although this is common); the presence of a British health worker renders it much easier. In that important respect, labour substitution encourages both absenteeism and moonlighting. On the other hand, if a deploying organisation takes the (correct) view that permitting volunteers to work on their own in such high-risk situations is in breach of our duty of care, and fails to contribute to capacity-building objectives, then facilities in real need of additional human resource will be denied it. And, sadly this was the decision the SVP was forced to take in Wakiso District Uganda after over 3 years of engagement and unsuccessful dialogue with the District Health Office. In the absence of an understanding of the causes of low staffing, the very conspicuous absence of local staff effectively justifies and encourages gapfilling behaviour by volunteers. The Independent Risk Assessment commissioned for the SVP added further impetus to these concerns. Identifying lone working or 'unsupervised clinical activity' as a key element of 'unacceptable residual risk' in some Ugandan facilities, the Risk Assessment took an unequivocal position requiring that volunteers 'withdraw from undertaking clinical work in the absence of professional Ugandan peers, or should they become a substitute for Ugandan staffeven if this leaves the patient at risk' (Moore and Surgenor 2012: 20). At the time we were surprised to find that the Risk Assessment identified Mulago National Referral Hospital as presenting the most serious risk of lone working (Table 2.2) 8 : This hospital in the centre of Kampala is, of course, the facility with the highest number of healthcare workers and one of the very few facilities in Uganda employing specialists. 9 The Department of Obstetrics and Gynaecology at Mulago Hospital (in 2014) employed 47 specialists, 48 senior house officers, 100 interns (17 at a time on rotations) and 350 midwives. These figures may seem reasonable in a facility delivering 30,000 deliveries a year. However, how can the risk of lone working be so high in such a context?
The reality is that staff are often not present on the ground during their contracted hours and it is very rare indeed to see any specialists present on wards; they are conspicuous by their absence.
A study by a local clinician on the 'Decision-Operation-Interval' examined the time that lapses between the decision to perform an emergency caesarean and the operation taking place and the causes and effects of those delays. Whilst lack of theatre space emerged as the dominant factor delaying operations, the report also identified a whole range of 'personnel factors' (shift change-over delays, absenteeism or late coming) underlying delays (Figure 2.2): There is no scope in this book to discuss the consequences of low and unpredictable remuneration in any detail. Salaries are certainly below subsistence level requiring health workers to undertake additional work to make ends meet. The absenteeism that we witness is not a symptom of laziness or general demotivation; the more senior staff are typically very highly motivated and work very intensively deploying a high level of skill. But the overwhelming majority of this work takes place on a private basis. They are 'otherwise engaged' but often working long days and through the night with private patients and in private clinics or, in some cases, on NGO-funded projects. Shrum et al. had a similar experience in a project concerned with the installation of Internet communication systems in Ghana. Here, key players frequently failed to 'show up' for work. The authors make the subtle observation that, 'It's not that anyone was trying to do anything except their job. . . . It's that they have a lot of jobs' and were constantly engaged in trying to make money (2010: 160).
Absenteeism and moonlighting present specific challenges for programmes, such as the SVP, committed to avoiding labour substitution wherever possible. Put simply, where Ugandan staff are regularly absent and the risk of lone working is high, we are unable to place professional volunteers   The following section considers the role of professional volunteers from a more informed human resource perspective, arguing strongly that volunteer deployment must be framed and negotiated within an evidence-based understanding of local human resource dynamics. In so doing, it also emphasises the importance of multi-disciplinary expertise and not leaving these kinds of decisions to individual clinicians who may arrive in an LMIC with little understanding of human resource management in low-resource settings or even of international development.
The title of this chapter 'First do no Harm' 10 is taken from the Hippocratic Oathan ethical statement governing the conduct of the medical profession. At face value, the Oath and its interpretation through the General Medical Council's 'Good Medical Practice' Guide (2015) do not suggest any major contradictions or tensions for doctors. Put simply, it requires doctors to pledge to put the needs of patients first and 'do no harm' to them. An earlier version of the GMC guide included a paragraph stating, 'Our first duty is to our patients, not to the Trust, the NHS or to Society' (2012). This implies a prioritisation of the one-to-one doctor-patient relationshipa highly individualistic approach to patient well-being which guards against political and pecuniary interference. However, it fails to grasp the potential unintended consequences of this approach when doctors are working as 'outsiders' in a foreign health system. 11 Hurwitz suggests that this simple message masks greater moral complexity in the face of 'bizarre moral predicaments' as 'new obligations thrust on doctors may conflict with their first responsibility to care for patients' (1997: 2). Although Hurwitz refers to the challenges of working in 'extreme circumstances', there is no explicit reference here to diverse international contexts. The updated (2015) version simply states: 'Make the care of your patients your first concern' (p. 0) potentially opening up opportunities for a more holistic interpretation.
The prioritisation of the doctor-patient relationship is often evident in the motivations expressed by professional volunteers applying for international placements through comments such as 'wanting to help people' or 'make a difference'. Many of the professional volunteers motivated to work in LMICs are motivated not only by clinical concerns but also by religious convictions. And these 'Good Samaritan' motivations often accentuate the desire to focus on individual patients rather than understanding and responding to systems. 12 Furthermore, whilst many professional volunteersand especially those with prior experience in low-resource settingsarticulate an interest in sustainability and longer-term change, they rarely interpret this as challenging their immediate commitment to individual patients. In other words, that systems change and immediate patient care may lie in some tension.

Similarly, respondents in an evaluation of the International Health Links Scheme (Ackers and Porter 2011) expressed concerns about UK volunteers being left to work in the absence of supervision:
We should say that we wouldn't send over junior British staff unless there's a senior [local clinician] on the wards and I wonder if that might set a bit of an example.
The SVP evaluation is peppered with similar experiences. In one example, a very experienced professional (short-term) volunteer described in his postreturn report how, as soon as he arrived on the ward, the local consultant made an excuse that his partner was not feeling well and leftand then failed to return. The consultant in this case explained how, in the time frame of his short (10-day) stay, he managed to clear the backlog of untreated oncology patients and relieve congestion. Clearly, the patients were direct and immediate beneficiaries of this process but it would be impossible to justify this kind of voluntarism from the perspective of skills exchange or sustainability. And as soon as the volunteer returned to the UK, the wards would rapidly re-congest. Indeed, a more impactful response generating greater patient benefit in the long term may have been to reply 'I'm sorry but if you go I have no choice but to do the same'. This is the culture that we have been trying to embed within SVP relationships with an increasing emphasis on conditionality as relationships mature and mutual understanding grows.
In another quite different situation, the arrival of a group of American midwifery students at a Ugandan health centre was marked by staff absence. It is hard to say in this case if the arrival of foreign students encouraged staff to absent themselvesbut they were certainly not planning to welcome them and the SVP obstetrician noted that the level of absenteeism was unusually high: The Americans have been covering up a shocking lack of staff at [facility] in the last two weeks which is good for the women but is making me grind my teeth. Essentially it seems that most of the staff have been individually summoned for trainings of various kinds by various agencies without any co-ordination with the sister or doctor in charge at the facility leaving us for days at a time without a neonatal nurse (V).
One of the most tangible signs of labour substitution is the placing of professional volunteers on staff rotas. And however much we discuss with the local partner, the problems with this is it remains a high expectation whether the visitors are consultants or students. We were aware of these tensions before the start of the SVP and issued clear guidance to all parties that professional volunteers should not be placed on staff rotas except in exceptional circumstances. 14 Quite understandably, local health workers are often upset about this and resent it, expecting volunteers to relieve them of very burdensome tasks. This reflects misunderstandings about the role of volunteers (and of Health Partnerships and AID more generally) accentuated by years of experience of missionary-style labour substitution voluntarism. Some local health workers will challenge the decision not to permit volunteers to go on rotas, suggesting that volunteers are work-shy voluntourists and more interested in going on safaris than supporting them. And this may well reflect their experience of volunteers. Challenging this culture of volunteering has proved a challenge within the SVP but we are confident that consistency in response is essential. The following Ugandan clinician who was part of a focus group argues forcefully against allowing volunteers to go on staff rotas on the basis that this will undermine co-working and encourage absenteeism: I don't support the idea that they go on the rota. I would not support thatthey will leave all the work to her (the volunteer). I've seen it. Once you add someone extra on the rota someone in that group will disappear for a year as long as they know the volunteer is there (FG).
Whilst this expectation was almost always experienced at the start of placements, it is by no means only at this stage. For most professional volunteers, it is an ongoing process involving complex negotiations at many levels. In one case, a volunteer who spent over a year in Uganda was constantly under pressure not only from her peers but also from the hospital superintendent (in this case, a British volunteer himself) to become involved in routine service delivery and be placed on local staff rotas. 15 She battled on a daily basis to resist service-delivery roles for over a year. Sadly, when she returned to Uganda after some months in the UK she immediately found that the expectation had increased. Staff assumed, as she knew the place and had experience of working there, she could immediately substitute for local health workers. In her monthly report she identified the 'main obstacles to achieving her objectives' as follows: It's just that I seem to be left to do things on my own now a lot. Frequently I am doing the ward round alone with or without the intern as the only other midwife on the ward is in the Waiting Home for half the morning. Because I have been here so long the midwives treat me as one of the rota staff, which is lovely as they accept me and trust me, but means I can't do admin and prep for teaching as they assume I am always going to be there to do the ward round. And as there is often literally no-one else to do it I can't really just disappear to do teaching prep etc. so my objectives changed -I think that is probably a natural progression in this type of work after one has been there for a while (V). This case has encouraged us to reflect on another deeply held assumption within the international volunteer deployment community and among hosts that long stays are far more valuable in terms of development impact. The issue of length of stay is discussed in some detail in Ackers (2013). What is clear from the experience of this volunteer is that the presumption of gapfilling increased with length of stay and became very difficult (impossible) to negotiate as time went on: It would seem offensive now to the staff who I have got to know so well and so closely if I were to stop working the moment there was no-one to work with. This situation may reflect a failure on the part of local staff to understand the role of professional volunteers, which may itself reflect a failure on the part of the deploying organisation, the host management team or the volunteer themselves to understand capacity-building approaches to international development. In many respects, we are dealing here with trying to effect in-depth ongoing culture change in an environment in which many of the actors involved either don't understand or don't subscribe to that (systems-focused) approach. One midwifery volunteer describes her experiences: On my first day all the midwives left to have their lunch. I was the only midwife on the ward of 27 labouring or newly delivered women. I think there will always be difference in opinion as to whether we are replacement labour or not (V).
This presents serious challenges when placing professional volunteers in the Ugandan healthcare system where the lack of senior staff or their failure to be present on the wards leaves more junior staff and students in situations where they have to work on their own and outwit the bounds of their competency. Lone working without supervision is normalised for Ugandan healthcare staff and it is unsurprising within this culture that volunteers are expected to do the same. One UK consultant clinician explained in her report how senior staff 'walked off the ward' the moment she arrived. These are common (normal) experiences in Uganda. The following excerpt from a blog written by an LMP obstetric volunteer working in a facility delivering 30,000 babies a year (over 80 a day) illustrates the problem in more detail: The 2 weeks leading up to Christmas were the most intense weeks that I've had at [the hospital]. All of the Senior House Officers [clinical trainees] were on exam leave and to make matters worse the interns [junior doctors] were on strike because they hadn't been paid. I was the only junior doctor on the rota to cover labour ward, theatre and admissions (there would normally be 3-4 SHO's and 4 interns)! Two seniors [specialists] were supposed to be covering labour ward during the exam period, however often only one would turn up and go to theatre leaving me alone. One day no specialists turned up at all, so I wasn't able to open theatre when there were 8 women waiting for caesareans. A woman presented with cord prolapse so I had to take her to theatre but she was the only caesarean that got done. To say I felt vulnerable would be an understatement, and in true [hospital] style everything you could imagine happened: eclampsia, twins, breech deliveries, abruptions, ruptured uteri. One particular incident happened when I was alone in admissions. A woman arrived in a semi-conscious state following an eclamptic seizure, and was having an abruption (premature separation of the placenta leading to heavy vaginal bleeding). It was very hard to auscultate a fetal heart beat and I feared the baby was dead. After delivering the baby with a vacuum it needed urgent resuscitation. I attempted to resuscitate the baby but it was futile, I didn't have a towel to dry the baby and the resuscitation equipment was broken. A very frustrating and upsetting day (V).
This volunteer was deployed via the LMP in the year prior to the SVP and her experience had a profound impact on project design. During that time, a HUB partner working in Gulu Regional Referral Hospital recounted the experience of a volunteer midwife who, initially put herself on the staff rota. However, the local midwives stopped coming in because they thought, 'Oh she is there so that's OK'. So she took herself off the rota and started to come in at different times and did an assessment and made decisions about where her work was best needed. So she wasn't on the rota because, especially when it came to the evenings, she was invariably the only midwife there. I had a long chat with some other doctors and they said they'd seen the same thing. Two young [volunteer] doctors turned up and all the senior staff went on holiday the next day and that's unacceptable. It's very difficult to extract yourself from that situation.
The case illustrates the relationship between lone-working and competency with early-career volunteers often under serious pressure to perform tasks that fall outside their experience and confidence.
This situation is by no means limited to obstetrics and gynaecology. This is just the department we are most familiar with. And as the SVP began to recruit and place anaesthetists we became acutely aware of similar problems. SVP anaesthetic volunteers were being repeatedly put under pressure to open theatres on their own due to a lack of local specialists. This came as something of a surprise as Mulago was one of only three hospitals in Uganda with specialist anaesthetists, most of whom have been trained with support from the Association of Anaesthetists of Great Britain and Ireland (AAGBI) and partner organisations in the USA and Canada. The reality is that there is no shortage of specialist anaesthetists in Mulago. However, they are rarely present to fulfil their local public duties or to work alongside professional volunteers. The initial advice from the AAGBI was that we should only place anaesthetic volunteers in Mulago, Mbarara or Mbale where UK-trained anaesthesiologist were in post. One consultant anaesthetist volunteer spent her first 2 weeks in Mulago and reported on her early experiences following initial meetings with local staff: As far as my activity in Mulago [I plan to have] a non-clinical role as my working hours coincide with the presence of skilled and experienced anaesthetic staff.
Several weeks later, her perspective shifted when the reality of working in the National Referral Hospital became clear: Staff absences and late starts are endemic and my presence alleviates the situation at times. As I have spent more time in Mulago I have got caught up in service provision. I'm feeling stressed, exhausted and like I'm failing on every front. The obstetric anaesthesia lead is rarely in labour theatre. There are always local practitioners (anaesthetic assistants) when I'm working but there has been 1 episode of me being the most senior anaesthetist on the floor with 3 Ugandan students for me to supervise. The senior Anaesthetic Officer (whom I contacted) who was supposed to be present felt no unease with the situation. The students' neonatal resus skills are not yet well established and I felt the whole setup left both me and the students exposed. The cases were of prolonged and obstructed labours and both mothers and babies were at high risk of complications.
There As a result of this feedback and the volunteer being put in a situation where she had to open up theatre on her own, we requested that she work in other facilities. Similar experiences were had by anaesthetic volunteers placed in Mbale where the specialist worked almost all of his time in the private facility. Mbarara was a significant and unique exception. The consultant anaesthetist in Mbarara embraced the logic behind the co-presence principle before we even used the term issuing instructions to his staff that they must remain in the workplace until the UK volunteer herself left. This placement had proved one of the most successful with clear signs of sustained improvement many years after the volunteer left due in large part to the attitude of the local mentor.
The final case presented here took place during the professional Risk Assessment process and was picked up by the risk assessors in their report (Moore and Surgenor 2012: para36): 36. As a condition of ethical approval by the Hospital Ethics Committee, we were told that medical students were required to work during the weekends and at night. Both the volunteer and medical students spoke about difficulties accessing senior medical colleagues during the night. We were informed of a particular night shift wherein there were 2 still births, a death on the Maternity HDU and an obstructed labourobstetric and midwifery staff apparently refused to attend and assist because they were sleeping (which we were told is normal practise and they are not to be disturbed whilst sleeping). We understand it was left to volunteers to work through the problems as best they could. Medical students explained how they were often goaded into carrying out clinical examination or diagnostic procedures they did not feel competent to perform, and whilst they declined to carry out the procedures, they explained how this created some tension with Ugandan medical students also working at the Hospital. We were concerned here about the level of clinical supervision and support, but also the security implications of working at night. This case was also reported to us by the volunteer, resulting in a formal complaint and the promise, on the part of the Ugandan facility, to investigate further. We were not aware that this took place. In fact, the British obstetrician did wake one of the sleeping Ugandan doctors who then refused to assist her and complained at being woken up. The British doctor reported this situation in the patient's medical notes precipitating angry exchanges as Ugandan doctors pressurised her to remove the comments. This incident took place in the final 2 weeks of a 12-month placement causing serious anxiety for the volunteer. And, the pressure to undertake data collection during the night (on the part of the British medical students) came from their UK obstetrician supervisor keen to gain round-the-clock data collection for his research paper. When we contacted the obstetrician about this he responded defensively expressing the view that 'clinical' mentoring should and could be distinguished from risk assessment. In other words, risk was not his problem: Risk assessments are really issues for [sending organisations] rather than clinical mentors and I would not like to [get involved].
Sadly, service-delivery roles are also a direct response to the demands of foreign visitors, often keen to gain access to patients and conditions that they are unable to achieve at home. One of the worst examples of service delivery we have witnessed in Ugandain this case entirely focused on training American doctorsis described by an SVP volunteer: The Americans have kind of taken over (one of the obstetric) theatres. They have got some senior residents in special training and they have got these really junior doctors who are increasing their caesarean section skills. They have been here for a month just doing a lot of sections. They work during the day shift.
Interviewer: So their objective is to train the US junior doctors and they take up the whole theatre? Are there any Ugandans in there then?
No, I think they have been doing this for several years they have got introduced to everybody in one of the morning meetings and one of the guys said we have been coming here for six years.
Interviewer: So, you think the main point is to train the American junior doctors because they cannot get that access over there (in the US)?
Yeah (V) This situation is entirely unacceptable and unethicaleven if it did mean that Ugandan mothers were being treated for free during that period with US equipment and staff. Not only does this type of intervention undermine the Ugandan health system, but it also caused problems for SVP volunteers attempting to achieve a level of co-working with local staff. 16 The following paediatrician contemplating applying to the SVP describes her experiences of volunteering as a medical student and her concerns that these forms of gap-filling voluntarism generate dependency: I'm not sure whether to go again. I first went to Uganda in 1985 as a medical student to a mission hospital. All the doctors and nurses there were ex patriates. They had their fingers in the dyke really. Although the medical superintendent was Ugandan and they did a great job looking after patients when they were there, there was no succession planning. There was complete dependency on the foreign staff. I guess it was a mission hospital model (V)

CO-PRESENCE AND KNOWLEDGE BROKERAGE
The previous section has discussed the risks and unintended consequences of labour substitution models of volunteering. Chapter 1 described THET's mission in terms of 'leveraging the knowledge and expertise of UK volunteers to build human resource capacity'. Clearly, deploying volunteers to replace local staff does not begin to operationalise that goal. The emphasis on knowledge in THET's mission could, arguably, be achieved through other forms of intervention such as donating books, providing on-line training or increasing training opportunities in the UK. It goes without saying that British health workers represent an important resource. They possess valuable knowledge gained through undergraduate education and subsequent continuing professional training and experiential learning. Of course, this is a diverse population and their skills, knowledge and personalities will vary widely. The fundamental question for projects such as the SVP is how can this resource (i.e. the embodied knowledge of UK health workers) be mobilised and deployed to offer optimal benefit to the Ugandan public health system? And what added value does flying them out to LMICs (human mobility) bring?
Our familiarity with the research on highly skilled migration and knowledge mobilisation made us aware of the complexity of knowledge itself and how difficult it is to simply 'move' it from one context to another and expect it to stimulate innovation or behaviour change. Although we are aware how complex these debates are, it is useful to summarise them here if only to help us understand what we mean by 'knowledge' in the Health Partnership context. 17 Williams and Balaz (2008b) distinguish various types of knowledge suggesting that some forms of more explicit knowledge (such as technical skills) may be transferred internationally via text or virtual means. He contrasts this with 'embodied' knowledge where learning takes place through doing, is highly context-bound and requires greater co-presence (or face-to-face interaction 18 ) and stronger relationships. Meusburger similarly identifies a 'missing distinction' in debates around the spatial mobility of knowledge, between knowledge and 'routine information' suggesting that, 'codified routine knowledge that can be stored in databases has to be distinguished from intuition, foresight and competence based on years of experience and learning' (2009: 30).
Whilst it is useful to identify explicit and tacit knowledge as opposite poles along a continuum, in practice, the categories are fluid (Meusburger 2009: 31). And the distinction begins to lose its significance when it comes to the application of knowledge. The capacity-building and systems change objectives of Health Partnerships demand highly complex forms of knowledge transfer, combining technical skills with mechanisms for their translation into socially relevant outcomes. In that sense, even much standardised forms of knowledge (clinical skills) need to be complemented with highly contextualised knowledge to support effective implementation. As Williams notes, while it is important to distinguish different types of knowledge, 'one of the keys to their valorisation is how they are combined' (2006: 592).
Williams and Balatz's paper on knowledge transfer in the case of returning Slovakian doctors opens with the assertion that, whilst health worker migration is an 'inescapable feature of the health sector . . . there has been relatively little research on mobility as a conduit for learning and knowledge transfer ' (2008a: 1924). The paper identifies a range of knowledge acquired by doctors including 'technical skills, academic knowledge, cultural knowledge, management know-how and administrative skills' (p. 1925). They suggest that whilst some knowledge may be transferred electronically perhaps through reading and published protocols, other forms of 'embodied knowledge' are 'rooted in specific contexts, physical presence and sensory information and may include participation in clinical practice'. And these forms of knowledge are 'grounded in relationships between individuals' and in socialisation processes. The successful application of knowledge combinations, according to Williams and Balatz, requires 'co-presence ' (2008a: 1925). The authors describe the opportunities for actors in this knowledge exchange process to act as 'boundary spanners' operating in places of 'unusual learning' where perspectives meet. And the conditions for this higher level of comprehensive knowledge exchange are not simply met by crossing national or other boundaries but by the quality of relationships at those boundaries (p. 1926). Meusburger contends that understanding the 'spatial mobility of knowledge' demands awareness of communication processes (2013: 29). Even where levels of explicit knowledge/skills are deemed higher in the UK, complex communication and strong relationships are required in order to contextualise that knowledge and translate it into effective practice in a Ugandan healthcare facility.
Meusburger is quite right to identify a range of 'assumptions' that shape the quality of relationships, including the impact of asymmetric power and the importance of non-verbal communication emphasising the importance of co-presence or 'F2F' contact. He also usefully distinguishes the types of individuals involved on the basis that knowledge may move differently between different kinds of stakeholders and practitioners and identifies a number of factors influencing relationships and communication process.
These include the 'cognitive abilities, ideology, interests, motivation, attention, emotions, and prejudices of the recipients and the milieu they are embedded in' (2013: 33). The emphasis on communication here is essential but in the context of multi-lateral exchanges. And participants in this co-learning process will bring different forms of knowledge to the table.
In order to achieve the goals identified earlierwith a strong focus on co-learning to support systems change -Health Partnerships need to focus on identifying mechanisms to facilitate the kinds of relationshipbuilding conducive to behavioural change. Co-presence is a necessary pre-requisite for the kinds of relationship formation conducive to knowledge translation.

The Sustainable Volunteering Project and the 'Co-Presence' Principle
Our experience of the risks associated with labour substitution or 'locumvolunteering' coupled with our research on knowledge mobilisation (albeit in a rather different context of scientific mobility) encouraged us to import 'co-presence' as a core operational principle shaping volunteer deployment in the SVP. 19 In this context, the doctor (or health worker) as a professional volunteer becomes a knowledge intermediary first and foremost rather than a 'carer'.
In practical terms, 'co-presence' simply means that UK professional volunteers should always be physically working alongside Ugandan peers in an environment that promotes opportunities for knowledge exchange. Co-presence does not imply that professional volunteers do not engage in clinical work. However, when they do so they must be appropriately mentored and engaged in active mentoring (according to their needs and the context). Co-presence is a composite concept representing the quality of relationships. Effective relationships play a number of distinct but related functions in the context of professional voluntarism. These include: • The promotion of volunteer safety and mitigation of risk (discouraging lone working and ensuring compliance with competency principles). • The facilitation of effective knowledge transfer (through training, mentoring and co-working). • The process of embedding reciprocity, accountability and conditionality.
Implementing co-presence has been and continues to be a challenging process. It has met with resistance from not only local Ugandan staff (as noted earlier) but also some volunteers keen to optimise their opportunities for clinical exposure and often frustrated at the inability to intervene when local staff are absent. Nevertheless, we believe that is has begun to be understood and recognised as one of the features of the SVP. From an operational and evaluation perspective, it is implemented through a monthly reporting system which requires volunteers to state whether they have been able to comply with the principle and identify situations where the project managers need to intervene. This has been reinforced through regular interviews with volunteers and their hosts, site visits and bi-annual workshops. Co-presence now forms a core component of any Memoranda of Understanding governing relationships within the SVP and is increasingly subject to more concerted conditionality requirements. In more recent work it has shaped volunteer engagement in degree-level teaching and the functioning of the Ethical Electives Project (Ahmed et al. 2016b).

SUMMARY
Following the discussion of objectives in Chapter 1, this chapter has outlined the dynamics of the human resource environment within which capacitybuilding projects, such as the SVP, deploy professional volunteers. The SVP, in common with most volunteering schemes, has faced the multiple dilemmas of attempting to place professional volunteers in contexts, often at the requests of senior managers, only to find them left to work on their own in high-risk and challenging service-delivery roles. Not only will volunteers find that many of the staff employed to work in these facilities are not routinely there but their very presence, as volunteers, will encourage others to absent themselves. And volunteers themselves (particularly doctors) perhaps motivated by ethical principles to respond immediately and unquestioningly to patient needs or, more commonly, by their own desire for clinical immersion and the opportunities to practice on complex cases, often enjoy and seek out such high-risk 'Ninja' 20 medicine. Enforcement of co-presence is essential to change the culture of volunteering and the systems damage caused by passive and dependency-generating gap-filling. In that respect, co-presence must avoid becoming one of the conditionality principles that Moyo suggests have 'failed miserably' to constrain corruption and bad government because they were 'blatantly ignored and AID continued to flow' (2009: 39). Conceptualising professional volunteers as knowledge intermediaries in systems change interventions places a firm emphasis on the co-presence principle. Co-presence cannot guarantee effective learning, but it is a pre-condition of it.