Health service resilience in Yobe state, Nigeria in the context of the Boko Haram insurgency: a systems dynamics analysis using group model building

Fifteen variables were confirmed by GMB participants as key in accounting for health
systems functioning during the crisis. These are considered, in turn, below. During
the course of modeling, an additional 12 variables were specified as intervening factors
clarifying linkages between variables, with staff commitment and motivation, the most
elaborated. The linkages identified between all these variables are shown in Fig. 1, the causal loop diagram developed by participants during the GMB exercise.

Fig. 1. Causal loop diagram of factors influencing health utilization and service quality
during the insurgency in Yobe developed through group model building

Insecurity

Between June 2011 and August 2014 there were numerous documented attacks by Boko Haram
within Yobe state resulting in an estimated 1341 fatalities 28]. Incidents peaked in October 2012, with over 12 reported attacks and 140 fatalities.
There were 252 deaths in Yobe state in the first 6 months of 2014 alone 29]. Interviews documented many incidents vividly illustrating this prevailing insecurity:
One health facility worker, for example, reported:

When Dogon Kuka was attacked, people ran away and the health facility was closed for
almost three months. One security guard was killed in his house.

Another health worker told of how one day gun shots started when she was in the office
and there was no way she could move to her house easily. All the staff lay on the
ground, weeping on the bare floor tiles.

Due to such insecurity, many people felt compelled to migrate out of their villages,
and health workers left their workstations for safer places. Insecurity also created
movement challenges, as people were afraid of being attacked when traveling because
insurgents often blocked the roads and attacked people.

During the GMB session one participant reported an incident regarding some health
workers who were attacked after the prevailing 6.00 pm curfew because they had been
called to render services at the hospital due to an emergency. Even though they had
IDs to allow them to travel during curfew hours, the Nigerian police had physically
harassed them. There were a number of reports of the police abusing their power through
assaults. This situation was aggravated when the ‘Escobar’
⁷
phenomenon was at its peak, since one could not differentiate between the official
security forces and criminal gangs because they were all wearing uniforms, carrying
guns, and driving HILUX vehicles. Such incidents were reported to have been at their
peak during 2012 and 2013.

Migration of people

As the bombings, abductions and killings became wider and more indiscriminate in their
scope, increasing numbers of people relocated to what were perceived to be safer areas.
When attacks were concentrated on urban settings, migration was to rural areas. When
urban centers such as Damaturu became more secure, attacks shifted to more rural areas,
and migration was towards the towns. For example, one of the interviewees in Damaturu
reported:

That time around, the patients have even migrated and left the town because it is
our catchment area, Damaturu, that was affected by the activities of the insurgents
most especially in the Gwange area.

A GMB participant expressed similar sentiments and stated that in the beginning the
attacks were concentrated in Damaturu and people were moving from Damaturu to various
places. However the situation changed when Damaturu was heavily secured by armed forces,
the insurgents turned to the rural areas, and the migration patterns reversed.

Migration of health workers

Health workers were not spared in the killings and threats. Indigenous workers generally
abandoned their workplace temporarily and returned whenever the situation calmed down.
However, most non-indigenous workers left permanently. Records gathered from the Yobe
state Human Resource Information System (HRIS) indicate that the most affected cadres
of staff were doctors and nurses/midwives as compared to other staff as presented
in Tables 1 and 2.

Table 1. Yobe state health workers by selected local government area (LGA)

Table 2. Yobe state health workers by hospital

With the exception of Gujba LGA, Table 1 indicates that the number of health workers in other cadres had increased in 2014
compared to 2010, which is an indication of a higher retention during the crisis.

However, figures in Table 2 indicate a significant decline in the number of doctors and professional nurses/midwives
across each of the three LGAs over the same period. For example, the general hospital
in Buni Yadi lost all its three doctors it had in 2010 and there was no doctor left
at the hospital by June 2014. The general hospital at Potiskum lost about 90 % of
its doctors over the same period. Only Sani Abatcha State Specialist hospital seemed
to have significantly fared better, losing only 30 % of its physicians. 2014 figures
are likely to reflect an upward trend from 2013 when the State Governor lifted an
embargo on recruitment in recognition of the acute shortages that had arisen (HRIS
data is not available for intervening years). GMB participants stated that more doctors
are arriving and picking up jobs in hospitals now, hence the increase.

The trend for professional nurses/midwives is similar. For instance, Sani Abatcha
Specialist Hospital lost 75 % of its professional nurses/midwives while the general
hospital in Potiskum lost about 70 % of this cadre. Only General Hospital Buni Yadi
fared better, losing only about 25 % of such staff. Midwives were especially affected
because most of them were from other states deployed through the Midwives Service
Scheme (MSS) and almost all of them left the state at some point.

Key informants confirmed that during the crisis, most health workers did not want
to stay in Yobe state. One participant at the GMB session  stated:

In the hostel, a lot of medical doctors left, the National Youth Service Corps (NYSCs)
do not come to Yobe at all, not even Damaturu. A lot of non-indigenes left, a lot
of indigenes who are outside Yobe couldn’t come. But now it has been reversed. There
were only four NYSCs who reported initially but now there are 16.

Participants reported that more NYSCs were currently coming to take up posts in Yobe
state.

Movement challenges/restrictions

The many dangerous roads where insurgents could mount ambushes presented significant
challenges regarding movement within the state. However, counter-insurgency measures
also led to major travel restrictions. Curfews were sometimes set to begin as early
as 4 pm, creating major logistical travel difficulties. At times the JTF blocked roads
that were considered too dangerous for people to travel on. Diversions, significantly
adding to journey times, would generally be provided, but on occasions no alternative
routes were available, preventing all travel. Although some categories of health staff
were provided with IDs to ease movements, there were still some challenges. One participant
noted:

The ‘Road Pass’ helps most times but still sometimes it wouldn’t because it seems
one is exposing himself and you don’t know who else is watching, even the security.
It may come across like a particular person thinks he is special and they may follow
you and attack you. So at times you may have the pass and may not want to use it.

Another participant added to this experience stating:

Sometimes you have to bring your stethoscope out of your car. Some of us have a ‘Road
Pass’ to enable us pass the queue but it does not prevent a search. When there is
high tension, when you go and meet the security officer, they don’t understand but
some of them that understand may agree to search you and let you go.

Supplies of drugs and consumables

GMB participants confirmed the view emerging from stakeholder interviews that availability
of drugs and other consumables at facilities was, in general, not severely disrupted
by the insurgency. The State Ministry of Health and facility managers minimized stock
outs by changing deliveries from the usual 1-month supply to a 3-month supply.

Additionally, facilities were encouraged to order new stocks when they had 50 % of
the drug stocks remaining. In the context of mass casualty events, some short-term
drug stock-outs were reported, but restocking was secured swiftly. The decentralized
nature of the drug supply chain, with drugs managed at the state rather than federal
level, made it much easier to transport drugs to and from the Damaturu drug store
whenever transport routes were secure. Flexibility in local supply routes helped secure
supplies when circumstances were volatile.

The GMB participants and key informant interviewees both reported drugs being brought
somewhere close to a facility for collection by facility personnel once access was
possible. There was some evidence of commercial drug supplies being more severely
impacted due to transport difficulties and loss of consumers’ purchasing power. One
local drug vendor reported that to get a lorry to take their drugs to Damaturu sometimes
took up to 2 weeks. Even when they got the transport, the price was double because
the drivers didn’t want to go to Damaturu due to the insecurity. Another supplier
reported:

I used to sell an average of 30,000 Naira per day but with the insurgency, I could
only sell about 10,000 Naira a whole week.

Physical access to health services

Physical accessibility of health facilities was one of the most important factors
influencing patient utilization of health services. Interviewed patients stated that
during the critical periods of the insurgency, it was extremely difficult to get to
a health facility due to travel restrictions and fear of being attacked. One patient
noted:

As you know, we can’t come to hospital easily because of their activities…people have
suffered a lot because of the insecurity, even if you are not well and you want to
come to the hospital it is a big problem…if you intend to go to the hospital you are
thinking of what is going to happen to you on the way.

Some pregnant women were reported to have died as a result of complications during
delivery due to failure or delay in accessing health facility care. One of the participants
in the GMB session testified about his close friend whose wife died in labour at home
because she could not be taken to the hospital due to travel challenges.

At some point, arrangements were made with the JTF for security personnel to collect
patients from their homes and escort them to the health facilities when contacted
through a medical emergency telephone number. Unfortunately, this service had later
to be discontinued after insurgents feigned illness and then ambushed the military
forces coming to attend them. This had resulted in an arrangement where those telephoning
the emergency number and providing a description of their vehicle were allowed to
use their own cars to access a health facility during curfew hours. The JTF then monitored
the passage of the vehicle from a position of cover.

Financial barriers

Movement restrictions have limited people’s access to gainful activities, hampering
their financial capacities. People could not trade their commodities for fear of attacks.
From the early months of the insurgency, farming of certain crops such as maize was
banned so that insurgents could not use those fields as cover to launch their attacks.
Also, due to curfews and direct attacks, banks were only open for a short period of
time and some were completely closed. There were also restrictions on interbank transfers.
Even those who had money in their bank accounts could not easily access this money
to transact business.

A number of interviewees noted that the financial barriers to service access during
the insurgency were, however, significantly less than they would have been without
the introduction in 2009 of a state drug subsidy scheme. This ensured that all pregnant
women and children under-five had access to free drugs. All drugs and materials required
to treat a medical emergency, such as a shooting or bombing, were also covered. These
arrangements, and the separate ‘free MNCH’ programme operating in some facilities,
thus moderated the health impact of the disruption of livelihoods resulting from the
insurgency.

Utilization of services

Interviews suggested that the level of security influenced the use of services in
a complex manner. If the number of attacks was high in a particular area, the utilization
of services was low because patients were afraid to travel to the health facilities.
A medical doctor from one of the most affected areas stated that most patients that
came to seek care did so when they were critically ill and required urgent medical
attention. Patients who were not severely sick would defer or decline attendance due
to insecurity. Mothers who had appointments for routine immunization, for example,
often missed their appointments.

However, if an attack involved large numbers of casualties then health facilities
nearby would be filled with patients injured during an attack. Areas that experienced
fewer or no attacks reported a general increase in patient attendance because people
had migrated to those places.

HMIS data on facility attendance confirmed great fluctuations due to insurgency activities
and related counter-insurgency measures. For example, Damaturu has at times been a
major focus of attacks, but through counter-insurgency measures has at other times
been seen as a relatively safe haven. Figs. 2 and 3 illustrate this pattern of varying patient attendance for the three LGAs for which
HMIS data was reviewed.

Fig. 2. Trends in facility deliveries: total facility deliveries per quarter for specified
LGAs. Source: Yobe State HMIS

Fig. 3. Trends in facility attendance: total facility attendance per quarter for specified
LGAs. Source: Yobe State HMIS

Figure 2 presents trends in total number of facility deliveries per quarter in three LGAs
from January 2012 to June 2014. The pattern indicates that whenever there was a decrease
in facility deliveries in Damaturu, there was a corresponding increase of facility
deliveries in Nguru. For example, during the quarter July to September 2012, the number
of facility attended deliveries in Damaturu declined to 21 deliveries from 797 in
the previous quarter, while in Nguru the numbers increased to 835 deliveries against
111 deliveries over the same period. By April to June 2013 Damaturu had experienced
a sharp increase in facility deliveries while in Nguru the numbers reduced to previous
levels. By late 2013 the pattern had reversed again, with deliveries in Damaturu steeply
declining and those in Nguru rising again. Attended deliveries in Gujba were relatively
stable, with some fluctuation, during this period.

Similarly, Fig. 3 shows total recorded facility attendance across LGAs from April 2013 to June 2014.
Reporting from facilities was not complete during this period, so reported rates are
assumed to be an underestimate of total attendance. However, with comparable reporting
rates across the three LGAs comparison of trends across the different settings is
likely valid.

This data suggests a marked increase in service utilization in Nguru during the course
of 2013 (in line with the increase in attended deliveries during this period). With
no reported insurgent attacks in Nguru, this appears to reflect internal migration
within the state to an area perceived as safe during a period of major insecurity
elsewhere.

Health worker workload

As most non-indigenous health workers left Yobe state, fewer health workers were left
to attend to patients. For example, as noted earlier, the number of nurses/midwives
at the specialist hospital in Damaturu dropped from 107 in 2010 before the insurgency
to 27 during the insurgency period in early 2014.

Increased patient numbers at times of inward migration or mass casualty incidents
exacerbated the impact of this on workload. For instance, through November and December
2011, Damaturu experienced a number of intense attacks where over 100 people were
killed, each resulting in many casualties requiring urgent treatment. With increased
workload the quality of services was compromised because the few available staff could
not effectively attend to all the patients. Interviews documented reports of patient
deaths in the hospital attributed to delays in treatment due to inadequate staffing
levels. One nurse noted:

I am a nurse by profession, so I have restrictions, I am not supposed to be clerking
patients, but when they meet me, I have to step out to do the function of a medical
officer and also nursing work…the pressure on me will definitely undermine efficiency.

Pressure on health infrastructure

Increased utilization of services at facilities also created pressure on health infrastructure.
Primary health care workers regularly signaled the potential impact of this on service
quality:

the quality of services is not at all what I could call standard because we have a
lot of crowdedness due to the turnout of the patients.

These pressures affected secondary and tertiary care also. Bed space was frequently
inadequate when patient attendance increased beyond planned health facility capacity.
Mortuary capacity was reported to have often been inadequate to meet demand following
mass casualty incidents. Hospitals had to ask the environmental department to assist
in clearing away the dead bodies. Pressure on health infrastructure was so severe
that it prompted the state governor to intervene to release funds to construct more
mortuaries and extend hospital wards.

Liaison with security services

Liaison with security forces was key during the crisis to enable essential movement
of health workers during curfew hours. Doctors were sometimes called to go to the
hospital and help with emergencies but due to curfews they would be restricted in
their movement. The Ministry of Health made arrangements with the JTF to allow health
workers to go to work during curfew hours through prior arrangement with local security
forces.

This was subsequently facilitated by the provision of identity cards to show to the
security personnel when crossing checkpoints. However, health workers reported sometimes
being reluctant to show IDs for fear that this exposed their identity to potential
insurgents who on occasions targeted killings of government personnel.

Community members also liaised with security forces whenever they needed to transport
a sick person to the hospital or, more generally, if they wanted to report information
about insurgents in their community. However, those seen talking to security forces
were considered a direct target of insurgents, with one participant in the GMB session
claiming:

If they see you with the JTF number in your cellphone, or you are talking to a security
officer, you are dead.

State officials would also regularly liaise with security forces to seek to reduce
curfew hours to enable citizens to move around more freely and go about their business.

Duty shift patterns and task shifting

The restrictions in movement due to curfews created challenges in the duty shift of
health workers who were scheduled to travel to work at restricted times. To resolve
this challenge, the Hospitals Management Board convened senior physicians and nurses
and made a formal change to the health worker duty shifts. Instead of three 8-hour
shifts, two 12-hour shifts were established. This ensured continuity of staffing at
facilities in a manner achievable with curfew restrictions.

Other less formal adjustments to working practices were also reported. In the context
of the crisis, health workers were reported to have assumed duties for which they
were not fully trained and licensed: a form of informal task-shifting. Such flexibility
in role allocations was widely perceived as a necessity in the context of the prevailing
crisis, and a move that has demonstrated the significant capability of less senior
cadres.

Financial support

As a result of the crisis the health sector required additional financial support
to address the increased demand for health services because of casualties. As noted
earlier, the State government was already funding free drug programmes for specific
groups before the insurgency, and this arrangement extended to covering the costs
of those attending for emergency intervention following insurgent attacks.

Given State responsibilities for funding contributions to the JTF security operation,
Yobe state budgets for all sectors, including health, were put under severe pressure.
Longer term development projects (e.g. health facility construction) were postponed,
but generally finance for the health sector was robustly supported. In 2013 the health
sector absorbed 97.3 % of its allocated budget, which represented a budgetary allocation
double of that allocated for the health sector in 2011 when the crisis had just begun
(and of which only 60 % had ultimately been disbursed).

Political will

Stakeholders attributed financial support to the ‘political will’ exhibited by the
State Governor. His visits to hospitals and other facilities in the aftermath of major
attacks were clearly appreciated. One participant at the GBM session noted

Actually, the renovation to some extent was triggered by the insurgency. The Governor
was not even happy about the capacity of the infrastructure and that is why he declared
a State of emergency on the health sector. I recall the Governor having been around
more than ten times, which is unprecedented. He comes at will, without being announced,
just to see everything that he needs to see and for some years back sometimes it was
hard to see the Governor.

Additionally the state provided security personnel in the cities and also facilitated
removal of some of the roadblocks to support movement. More specifically to the health
sector, on appreciating the loss of capacity in the health workforce with the loss
of non-indigenous staff, the state sanctioned the lifting of an established embargo
on employment to the health sector. Further, the state provided incentives to health
workers such as offering furnished houses and ensuring regular and on-time salary
payment, which was generally secured despite all the challenges with the banking system.

Community resources/cohesion

Community support was regarded as a key factor that helped community members to cope
during the crisis. This involved providing spiritual, emotional, and social support
to each other. Most of the interviewees mentioned that they survived by faith and
through prayers, indicating that religion had played a major role in holding the people
together, giving them hope to go on and endure the crisis and its suffering. Community
members organized transport for those who needed to access health facility services,
and provided shelter to those who had abandoned their homes due to insecurity. One
GMB participant recounted how he had moved his family to a safer area but, in the
context of general fear and uncertainty, had recognized the need to build the trust
of the local community by attending Friday prayers in that area.

Community members were a key source of information regarding the insurgents. They
would alert the health workers not to conduct activities, especially immunization
campaigns, whenever the insurgents were operating within, or close by, their local
area.

Co-ordination of immunization is done through community leaders because they are the
ones that know when there is threat of insurgency around them since these health workers
are with them, they work hand in hand with them.

The community leaders are usually the one’s in the center of co-ordination. They are
the ones who usually see [the insurgents] when they are passing. The health clinic
has nothing for now as a way of coordination.

We have tried to work hand in hand with all arms…we had a series of meetings with
emphasis on community engagement in all the health interventions. When you connect
very well with the community and traditional leaders, you will get adequate information
including security information on how best they can deliver services. Sometimes if
health facilities are closed health workers are in their houses, they can render services
within their community to the level that is feasible…..

Staff commitment and motivation

The preceding quotation highlights the importance of one variable, not initially identified
during thematic analysis, which participants at the GMB considered important for explanatory
purposes: staff commitment and motivation. Interviews and discussion indicated that
commitment was generally high among the health workers remaining in post through the
insurgency. Health workers reported encountering many challenges but this was usually
reported to have not deterred them from providing services to the patients, although
they often worked under fear. Most of the facility in-charge persons reported that
they used to encourage and motivate their staff to ensure their availability at the
facility and this helped the health facilities to cope during the crisis period. One
facility in-charge reported:

Since I am in-charge of the facility, I don’t think there was any time I did not come
to work and I always organized my staff to come to work.

Patients who were interviewed backed up these assertions. They also reported on the
level of commitment with which the health workers operated as stated by one patient:

Thank God the truth is that the health workers are trying their best to provide quality
services.

Another patient reported:

The quality of service is good always and we find the health workers there…before
the insurgency the number of health workers were more than now but they are trying,
even now.

There were reports of health worker commitment in the face of significant danger.
One immunization officer talked of when a team had been intercepted by Boko Haram:

Insurgents came and warned them to stop the work because it is evil. If not, they
will kill them next time they see them doing it. The women thanked them but continued
with their work after the[y]… had left. One method they adopted is hiding vaccine
carriers in polythene bags… putting the vaccine in between luggage so as to hide it
from anyone watching them.