Perceived causes of the PROLONEGD 2017 nurses strike
Most interviewees echoed the official reason for the strike in stating that nurses wanted the government to honor its promise of signing their CBA to increase their pay and improve working conditions. The frustration of not having the CBA signed and implemented was reportedly compounded by the Salaries and Remunerations Commission (which advises government on public officer’s remuneration and benefits) defining nurses as semi-skilled and thus only eligible for relatively low pay scales, and doctors having had their CBA implemented when they went on strike:
“So, people were demoralized, they were like there was discrimination: other cadres are important [while] others are not important; some non-professional, others are professional. So, with that people decided we need to be recognized, let us go on strike and never come back unless our CBA is signed.” Hospital Manager-03.
Poor working conditions, including shortages of drugs, commodities, equipment and staff, were also felt to have contributed to demoralization among nurses, with several managers noting that these challenges had been exacerbated through policy changes such as devolution, free maternity care policy, and abolition of outpatient user fees at PHCs:
“We say that maternity services are free, yet the same facilities are not facilitated with the commodities. So, there is a gap such that the outcome of the mother and the child is 50-50 so because of that we are human beings and sometimes it’s not good always to see that your patients are dying in your hands. You would rather not provide that service.” Sub-County Manager-06.
Previous county level agreements on for example promotions, re-designations into appropriate job groups, and training had only partially been implemented, and therefore failed to address the general discontent and unrest among health workers.
A lack of clarity and agreement between national and county levels on how to handle national-wide strikes was seen to have contributed to the prolonged length of the strike:
“The central government used to accuse the county government. The county government … blame the central government that they have not disbursed the money … so there was a lot of shifting blames.” Sub-County Manager-06.
The nurses’ strike took place during a fevered election campaign period for national president and members of parliament and county leadership. With the outcome of the August 2017 presidential elections disputed and later annulled by Kenya’s supreme court and fresh elections having to be organised, politics reportedly took precedence over service delivery. Both sides -nurses and the government– then also took a hard stance: the nurses wanted their CBA demands met before they resumed duties, while the government considered the nurses’ CBA demands too costly. To handle this impasse, the government gave ultimatums and described the strike as illegal, possibly breeding further resistance.
Some respondents - particularly community members - felt that an underlying issue was that those in government who should have been working to solve the issue did not prioritise it because they could easily afford private care; highlighting feelings of injustice and distrust. Notably, at the time of our interviews - conducted immediately after the strike - the nurses had not received all the agreed allowances, and their CBA had not been signed. Most respondents felt that we should
‘expect more strikes’, by nurses and other cadres.
“Yeah, we expect more strikes very soon because if you give one son [doctors] and deny another one [nurses] then expect more strikes very soon unless the government sits down and thinks about this issue …. They are not being given their rightful share as simple as that.” Hospital Manager-04.
Efforts to keep services running
In the face of major facility and service closures and disruptions, frontline managers enacted a range of strategies in their efforts to keep services open, and re-open others. Across these strategies we observe some positive relationships, alliances and effects across the system, but also some tensions and conflicts. Many of the strategies were piecemeal, inconsistent and difficult to sustain.
Prioritizing specific services and creatively using available health workers
At hospital level, emergency services were prioritized, but even these services, were difficult to sustain with depleted human resources:
“during the doctor’s strike she [the facility manager] had to make sure there were a few nurses, and that the nurses would actually keep the place running … during the nurses’ strike we were doing emergency services both medical and surgical. But it was very strained because we didn’t have enough nurses to be able to monitor these post-operative patients. So, at some point we had to stop doing operations unless it’s a life-threatening case, so we had to turn away mothers. It was very painful.” Hospital Manager-07.
During the nurses’ strike, staff in PHC facilities were keen to offer some outpatient chronic disease management services, particularly refills for HIV and TB patients, given the potential harms to patients and communities with treatment breaks. Middle level managers drew on available nurses in their efforts to offer priority services. Some were full-time nurses who opted not to strike, and others were probation or local contract nurses employed on terms that did not allow them to strike. Some of these staff were redeployed from PHC facilities to assist in managing emergency services at the hospitals.
Managers were also able to persuade some striking nurses to assist by appealing to their sense of duty:
“We were telling them...we understand the plight of the strike … [but] … at the end of the day you are striking; the people who are affected is just the local person. And it’s them [the local people] being the victims while the people you are fighting against they are not feeling it.” Sub-county Manager-01.
In one hospital (A) during the doctors’ strike, some Non-Government Organization (NGO) and privately practicing doctors were called upon to conduct emergency surgeries. The privately practicing doctors were engaged on locum basis (paid per number of hours/days worked), however, the arrangement did not last long as their payments were often delayed or missed out altogether. A similar strategy was employed during the nurses’ strike: nurses were brought in on a locum basis to assist in emergency services, but this arrangement only lasted for a few weeks (8-12 weeks) due to non-payment. More widely, some of the striking nurses were persuaded to assist with (NGO) sponsored health outreaches and services, by their managers. They were offered lunch and transport allowances to participate in these activities. These financial incentives were appreciated by nurses: they had their salaries stopped several months into the strike, and some were growing worried about their livelihoods in the face of no progress towards a strike resolution. Recognising the acute shortage of nursing staff, new short-term nursing staff were employed by the county government after several months.
Minimising and managing conflict
During the nurses’ strike, nurses who assisted through the strike reportedly faced threats and intimidation from striking colleagues, who felt that their power was reduced by colleagues working. This resulted in some nurses withdrawing their help, or offering services inconsistently or secretly.
“Some of us like me I was half strike half job. I would come peep and then I go. If I saw there were two or three patients that are very sick, I would treat them and run very fast so that the other nurses cannot find me here and beat me.” Peripheral Facility Manager-01.
Tension between striking and non-striking nurses led to striking nurses reportedly sending non-emergency cases to a relatively well-resourced hospital to overwhelm their working colleagues. The hospital nurse manager had to intervene and calm the situation by informing them that everyone has a right to strike or not to strike:
“… there was a division, those who were on, and those who were out. It was seen as a betrayal, the ones who were on duty it’s like they were betraying the others. So, I went to their solidarity corner and I told them it is your right to go on strike and it is their right not to go on strike, so everyone to play the ball in his own court and it calmed down and we continued offering services because before I addressed them when they saw each other they were abusing each other live [out loud]. Hospital Manager-02.
Drawing on NGO staff, other staff cadres and students
NGO employed staff did not seem to face the same pressures to join the healthcare workers’ strikes, and some NGOs employed additional staff to cover services such as HIV and TB in PHC facilities and hospitals. Interestingly, in one hospital, an NGO tried to pay striking nurses to work on locum basis but those nurses refused, possibly under pressure from their colleagues.
Given the challenges of finding adequate health workers to support services, there was some reliance across the system on support staff (staff not involved in direct patient care) and other cadres. For example, support staff in some PHC facilities were dispensing drugs for minor ailments and refills for TB and HIV clients during the nurses’ strike. In one subcounty, public health officers were instructed to offer immunization services (usually run by nurses). In one hospital (B), the nurse manager negotiated with a local colleague to let their nursing students remain at the hospital newborn unit under the supervision of the nursing in-charge. Also in hospital B, managers allowed clinical officers (non-physician authorized to perform routine medical duties and procedures) who had additional qualifications in reproductive health to conduct Caesarian Sections during the doctors’ strike. During the nurses’ strike, clinical officers were also allowed to take on postoperative nursing care under the oversight of the hospital nurse manager. However, over time there were concerns that the quality of care was being compromised and the obstetrics and gynecology consultant stopped the initiative.
Links and interactions with private facilities
The above strategies were inadequate to fill the gaps, and so middle level managers used creative strategies to support access to services through local private facilities. For example, all three hospitals developed an informal system whereby they would perform emergency caesarean sections at the public hospital and have patients taken to local private facilities for post-operative nursing care. A similar strategy had been employed during the doctors’ strike where doctors working in the private-for-profit and NGO sector performed emergency caesarean sections in public hospitals and the public-sector nurses provided post-operative care. This pattern was to ensure the more expensive operations were performed at the public facility, and the cheaper nursing care services at private facilities, and therefore that patients were protected from catastrophic costs.
Subcounty managers also increased their supervisory visits to private facilities with an emphasis more on advice for handling volumes of patients rather than being too strict on quality control. Further, in some sub-counties, the managers supplied family planning and vaccines to local private clinics to ensure the services continued.
Although managers considered interactions with private facilities as necessary to support patients, there were concerns raised about the quality of care offered (particularly given increased volumes of patients), and the costs to patients:
“We partnered now … we used the [private facilities] ...although now some started charging a lot of money [for family planning and vaccines they were getting freely]. We had to stop them again, renegotiate with them but we had no option, that was the only way to go to make sure that the services were continued.” Sub-County Manager-05.
Another challenge was developing formal agreements. The county had many outstanding debts with private facilities from the doctors’ strike. In fact, some interviewees even felt private facilities might have been overly hiking the costs of caesarean sections to recoup some of their previous losses.
Support and action from the public
It was noted that there was relatively little protest and action by community members to keep public services open. One reason might be that many community members reportedly supported the need for health workers to strike:
“Even me I support them (striking nurses). The thing is everyone should respect another person’s work. If your colleague feels that he is suffering and is asking for his right, he should be given according to his education level and his work... Should he continue to suffer because people are dying? He shouldn’t claim for his right because he took an oath? That is not possible, because that is your work place... that is where he gets his daily bread.” KEMRI Comunity Representative-FGD04.
However, the perceived inappropriate handling of the strike by government leaders who could better afford private care contributed to feelings of frustration and disillusionment in the community:
“So, the good thing is in such situations [strikes], they should take them [nurses] seriously, and they should sit down, talk and give results which are helpful. ..Here in Kenya there are some counties that didn’t go on strike at all because their [county] government served them [the nurses] the way they wanted.” KEMRI Community Representative-FGD04.
There were some community protests and actions. For instance, after the death of two pregnant women perceived to be due to difficulties in accessing affordable maternity care, some urban residents demanded that their facility provide maternity services at the minimum, and engaged the media to support their cause. They were successful in persuading public sector health workers to attend to pregnant women in the latter stages of labour. In this case the chairman of the local health facility, an elected community representative, apparently felt responsible to speak out and advocate for the rights of the poorest women in the community.
“As the chairman of this facility all these [people] are looking at me, now if I also will sit there and look down, that’s not right. There are some means and I’m still strong even if the strike was going on now. There’s a mother who lost her life somewhere here. [She was expecting] twins and she wanted emergency [care. Yet] we have experts here (facility)... we know they could be able to handle that situation. This is a special place, you know there are rich people here, poor, and there are the poor and the poorest. So we are looking at the poorest people to get [help]. .. and I’m sure that mother she wouldnt have died, but she died.” Facility Mangement Committee- FGD04.
In another example, community members reported protesting against a local private facility that denied maternity services to a woman in labour due to lack of funds, only for her to die at the facility’s premises while those who brought her were looking for funds and the facility to demand for payment before the body was removed. Community members reported that the facility was nearly burned down.
Discussion
Our exploration of the prolonged 2017 strikes in Kenya highlighted a wide range of negative experiences at both health system/service and community levels. This is despite considerable efforts by many middle level managers, and some community protests and activism, to keep some essential services running. Here we discuss the political and health systems context of the strikes, the reported effects of the strikes and the potential to learn from health managers challenges and initiatives in future. We conclude with recommendations based on our data to reduce the likelihood of such prolonged strikes and associated negative impacts in future.
The health worker strikes cannot be seen in isolation of the prevailing context. There were multiple interacting factors ranging from political changes, health system challenges and increasing unionization of health workers. The implementation of a new devolved government in Kenya in 2013 made public sector workers’ right to unionize more explicit, except for the police and army [
27,
57‐
59]. Meanwhile, longstanding issues including dissatisfaction with pay, working conditions and human resource management issues remained unresolved [
24,
25,
27,
30]. When doctors and nurses did go on strike, the doctors’ CBA was signed but the nurses’ CBA was not. Such background conditions, and in particular feelings of unfairness between cadres, has also triggered strikes in Ghana and Nigeria [
6,
60]. In Ghana, for example, strikes were called when doctors (but not other health workers) began being given additional duty hours allowances and – later – when the government decided to integrate these allowances into salaries (leading to higher pay scales for doctors than other health workers) [
6].
Health systems exhibit features of complex adaptive systems (CAS) which include self-organization, feedback loops and emergence [
61,
62]. Self-organization is a process in which the components of the system interact to create new states (termed emergence) which are often unpredictable or unintended [
62,
63]. As CAS, the health system has separate parts, but these are interconnected and governed by feedback, and almost always, a change in one part, has effects on another part [
62,
64]. Parallels can be drawn between the duty allowance saga in Ghana [
6] and the cycle of strikes culminating in the 2017 strikes in Kenya. In both contexts, the governments failed to recognize the complex adaptive nature of health systems [
61], by giving allowances to one group of doctors (Ghana) and by implementing only the doctors’ CBA (Kenya). Both the Ghanian and Kenyan governments adopted reactive responses, acting only when health workers went on strike as opposed to more proactively. In Kenya, county and national governments issued threats and ultimatums which did not solve the underlying problem. In both Ghana and Kenya, systematic analysis of possible reactions to interventions and consideration of context, power and interests of different actors may well have led to less dramatic impacts and shorter negotiation processes.
Also, important in the nature and length of the strikes, particularly the nurses’ strike, was the timing coinciding with national and local elections. Elections were already expected to be associated with unrest and to undermine the fragile public healthcare system [
65]. While the timing may have been a strategy intended to add pressure on the government to meet the nurses demands, in fact it led to national and county leaders being distracted from the strike and its’ effects on patient and public safety. Our findings suggest a wide range of negative experiences. Disruptions to services and reduced admissions have also been documented by other studies by our group: one documented that the strikes resulted in marked reductions in admissions with 4 out of 13 county hospitals having almost no admissions throughout the strikes another found that the nurses strike severely affected immunization services in government-run referral health facilities across the country [
27,
30]. Our finding of no obvious dip in outpatient service utilization during the doctors’ strike specifically is potentially linked to the presence of nurses and other cadres (such as clinical officers) in outpatients, but a forthcoming paper will characterize further the effect of both the nurses’ and doctors’ strikes on in-patient admission. Our interviewees highlighted the devastating effects of service disruption on staff morale and on households, particularly for the poorest households. Given that about 620,000 Kenyans are pushed below the national poverty line every year due to transport costs and health care payments even under ‘normal’ conditions [
33], the impoverishing effect of the strike for the poorest households is likely to have been enormous. As with other sudden shocks to the health system [
66], our findings support that the impoverishing effects of the strike are disproportionately felt by the poorest and most vulnerable
.
Beyond impoverishment, interviewees talked in dramatic terms about negative health-outcomes linked to the strikes, including deaths, with the poor again being the worst affected. A recent analysis of the effects of six previous nation-wide Kenyan strikes on mortality data in Kilifi County (before the 100 days doctors and the 150 days nurses strike) found a 75% increase in mortality among children aged 12–59 months during the strike period, but no change in overall mortality [
24]. The authors noted that the lack of change in overall mortality could have been because the strikes between 2010 and 2016 were relatively short, with only one lasting for more than a month (42 days). Evidence from other settings suggests that the effects of strikes on health outcomes are increased where emergency services are not available or the affected populations are not able to access viable (available and affordable) alternate healthcare services [
1,
3,
19,
67,
68]. In Kenya, the Irimu et al (2018) study reviewing admissions in 13 public hospitals during the 2017 doctors’ and nurses strikes noted that ‘preventable deaths likely occurred on a massive scale’, particularly for the poor [
27]. We identified similar perceptions in our study, but this may be in contrast with the more modest effects reported for prior strikes [
24] . Given that the Kenyan public health system has faced a series of shocks and stressors over the decades, additional research that can provide more detailed data on the impact of the prolonged strikes on mortality over time is important.
An ‘everyday resilience’ lens is relevant for analyzing the strategies adopted by managers in response to strikes, and for considering the impact of the prolonged strikes on the Kenyan health system. Everyday resilience can be defined as the ability of the system to maintain positive adjustment in the context of chronic shocks and stressors in ways that allow the organization to emerge from those conditions strengthened and more resourceful [
43]. Whether everyday resilience is observed and built in the face of chronic and acute stressors depends on the nature of the strategies enacted by health system actors, and the capacities that they can draw upon. Absorptive strategies buffer the system from shocks and return the system to its state with little or no change in structure; adaptive strategies result in some limited adjustments in the system structure or processes; while transformative strategies result in significant functional or structural changes [
43,
69]. During the nurses’ strike in Kenya, we observed that middle level managers enacted a range of absorptive strategies in their efforts to keep services open, including mobilizing financial, infrastructural and human resources to support continuity of some essential services. Adaptive strategies included some reorganization of staff and services offered, but more significant functional or structural changes - transformative strategies - were not observed during the strike.
Across all the strategies observed, managers drew on their social networks and alliances to persuade and negotiate with various actors across the public health system to assist. They also demonstrated creativity in ways of working with others such as the local private facilities and NGOs. To keep key services running, managers drew on a long history of working together and coping with diverse everyday stressors in health service delivery [
28,
41,
43]. Their relationships – or the ‘intangible resources’ they were able to draw upon - were sometimes invaluable in helping them cope with the shock of the strike. However, there was little to suggest that the broader system was undergoing positive adjustment to minimise the likelihood of future strikes or build preparedness in the event of any such strikes. Thus, there is little evidence that everyday resilience was being built over the course of the strike. Indeed, tensions between health system actors, including conflicts between striking and non-striking nurses (as also observed in South Africa [
7], may have lasting negative implications for health system preparedness for and prevention of strikes. Our study did not include views from private facility health managers, but private facilities were frequently mentioned by community members and health managers as places where the public sought alternative care. A potential future research question might therefore be to examine if and how private providers can contribute to building resilience capacities that the health system can draw on in response to future strikes.
Study limitations
We recognize that this study was only conducted in one county out of the 47 Kenyan counties, that interviewee’s perceptions of the nurses’ and doctors’ strikes were often intertwined and that we did not interview union officials, private health care providers and frontline health workers. Nevertheless, the trustworthiness of our data and analyses are supported through collection of data through several methods (interviews, observations, document reviews and surveillance data), allowing for some triangulation [
68] of findings, member checking [
54,
70] with researchers and health managers, and the use of an embedded approach that enhances trust between researchers and health system actors and hence the trustworthiness of collected data. Our sustained engagement with a wide range of respondents at county level and more widely across Kenya, supports the validity of our findings and their wider relevance.
Conclusion and recommendations
The recurrence of health worker strikes and the prolonged nature of the 2017 strikes highlights the underlying frustration and unrest amongst public sector health workers in Kenya. There is an urgent need for national and county governments to appreciate the complex adaptive nature of health systems and adopt systematic monitoring of different components and proactive thinking around possible effects (positive and negative) of interventions and policies. County and national governments need to rebuild relationships with healthcare workers’ unions and include them in the development, introduction and implementation of policy decisions that impact on health workers [
27]. Careful consideration is needed to review the compensation packages of health workers to ensure fairness within and across cadres, and the creation of a conducive working environment to offer quality services. This is essential to improving staff morale and reduce the need to strike.
Chima noted that an ethical approach to resolving labour disputes would require both employer and employee to recognize their moral obligation to serve public interest, reducing the incidence and effects of strikes [
1]. An ethical approach to strikes is challenging to imagine in the context of political contest, constrained resources, great power imbalances between decision making actors and unclear public accountability. All parties should carefully consider the prevailing socio-economic and political circumstances when planning and managing a strike, including avoiding deliberately planning strikes around election times or in other times of predicted disruption and distress, particularly for the poor [
7]. To ensure mutually respectful dialogue between parties, discussions may need to be facilitated by formal independent arbitration processes.
Recognising that strikes remain a real possibility, there needs to be adequate planning and preparedness in advance of a potential crisis [
71]. Given their key intermediary roles, and the challenges they faced in the prolonged 2017 strikes, middle level managers should be better supported by managers higher up the system to design and implement effective and sustainable responses to sudden shocks, including strikes. Responses to shocks should not only seek to preserve core services but also to ensure that the poorest households and communities are protected from health-related and financial losses. This would support a move towards more ‘ethical’ strikes where at a minimum emergency and essential services are sustained throughout a strike, threats and intimidation of striking and non-striking health workers are minimized, demands by workers are reasonable, and governments respect and honor agreements.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.