This study provides valuable information about neurological services at the national referral hospital in Zambia. Consistent with prior studies from the region, seizure, headache and stroke are the most common diagnoses seen at the UTH outpatient neurology clinic [16, 17]. Stable patients in the top three diagnostic categories, requiring only medication refills, were determined to be the optimal target for a task-shifting intervention to reduce patient wait times. Programs targeting medication refills have been shown to improve adherence and decrease physician workload [18,19,20]. Figure 4 illustrates the recommended approach to the processes around the UTH outpatient neurology clinic, outlining a starting point for effective execution of the task-shifting program in neurology.
Neurology guidelines for upper and middle-income economies recommend individual patient management delivered by neurologists prescribing treatment, supported by imaging and high-technology laboratory work monitoring [21]. Unfortunately, this approach in Zambia and in other LMIC is often not feasible due to lack of neurologists, diagnostic investigations and effective treatments. In fact, the neurological exam remains the most useful diagnostic tool [22]. Our study found that long waiting times for patients to see a neurologist and scarce neurology workforce are adding to the current shortfall. In fact, the UTH outpatient neurology clinic has increased from 1 to 4 neurologists during the study period. In 2018, they experienced the highest patient volume, which likely resulted from the addition of the fourth neurologist. The demand for neurologists, however, continues to exceed supply. Unfortunately, these shortages are likely to be sustained. Thus, guidelines are not translating into clinical practice in these settings because it is not realistic [23]. Alternative strategies to improve access to care for patients with neurological care are urgently needed in LMIC.
Task-shifting to nurses and community health workers may benefit numerous components of the neurology care delivery in LMIC. In Mozambique, training non-physician clinicians in a task-shifted model of antiretroviral therapy rollout for persons living with HIV, compared to physician care, was able to improve health outcomes including viral suppression, toxicity and mortality [24, 25]. Building upon previous findings, we believe common outpatient conditions such as epilepsy, headache and stroke, may benefit from protocols that make it easier for less specialized healthcare workers to manage these conditions.
In this specific context, we recommend shifting clinical follow-ups and refills of chronic medications to nurses, pharmacists and EEG/EMG technicians. These less specialized healthcare workers, however, have had limited training in managing these disorders, so implementation would require an adequate planning and training. A task-sharing model could be adopted first followed by task-shifting as non-physician healthcare workers gained more experience and confidence with their new responsibilities. Ideally, community healthcare workers will be able to determine whether these patients are truly stable and then take accurate treatment decisions. To do so, it is necessary to create a clinical environment that is sufficiently regular and predictable (i.e., only performing clinical evaluations of the same conditions and patient population) because it gives nurses and community health workers the opportunity to learn these conditions through prolonged practice [26]. As the healthcare workers gain more confidence in deciding whether someone is stable or not, the task could be completely shifted to them to identify stable patients without routine discussion with the neurologist.
It is important to note that strengthening training without changes in systems such as clinical operations and procedures, may lead to missing unstable patients, and thus may prove insufficient to improve patient care and/or outcomes [27]. Conversely, we believe that to maximize the accuracy of clinical evaluations made by community healthcare workers, final neurological assessments and decisions should be supported by a checklist of the most common conditions seen at the UTH outpatient neurology clinic. Research has shown that the use of checklists improves clinical outcomes and involves both changes in systems and in the individual and clinical teams’ behavior [28]. The virtues of checklists are well documented in LMIC, where community healthcare workers were able to substantially improve their performances of measure practices with the use of a checklist, as well as improving patient morbidity rates [29].
Based on the Consolidated Framework for Implementation Research (CFIR) [30] creating tension for change is necessary and feasible through better data collection. We encouraged the UTH outpatient neurology clinic to include key performance indicators (kpi): new visit rate, new appointment lag time and no-show rate. These kpis are measurable values that demonstrate how effectively a clinic is achieving the neurology task-shifting goals and overall healthcare objectives. These measures will allow evaluation of this approach, not only for baseline but for ongoing monitoring of the task-shifting program, as well as garner support from leadership. With the creation of urgency, elevating the matter to an organizational aim, there will be an increased awareness among providers, and therefore a stronger culture to improve neurological care using differentiated models of care. Task-shifting can have a relatively quick effect on utilization and on new appointment lag time by increasing the capacity of the services. Hence, there will be a need for shorter evaluation cycles of this neurology task-shifting model, that focuses on process optimization, avoiding radical workflow and operational changes. Task-shifting interventions in HIV populations in Africa may be used as a benchmark for comparing efficiency in a potential task-shifting intervention in neurology in similar countries.
This study has several limitations. The current study only considered the view of patients and providers in a national referral hospital and not populations in more rural settings in Zambia. Since our study is limited in its geographic reach, the findings may therefore be biased towards this population’s experiences and preferences. However, process mapping is a method that can be adapted and used to identify gaps in other outpatient neurology clinics in LMICs, which in turn can help those clinics to tailor the use of differentiated models of care to meet their identified needs.
In addition, our data brought to light other issues in the clinic such as the low appointment bookings rate, which raises the question of what communication barriers patients are experiencing when trying to book appointments in advance. Unfortunately this information was not collected as it was out of the scope of the project. These issues warrant further consideration and may pose opportunities to better communicate with patients by reinforcing the need to schedule appointments in advance or allowing patients to schedule appointments with the use of mobile apps and other technologies. Finally, this study focuses on a clinic in which fully trained neurologists are available to evaluate new patients and support community healthcare workers participating in the task-shifting models. Our suggestions would likely not apply to less-resourced settings in which no neurologists were present.
In Zambia, as happens in other LMICs, the treatment gap of neurological disorders is due in part to a small or absent neurology workforce leading to long waiting times to meet the growing demand for their services. Prospective solutions to reduce lag times may include task-shifting neurological care to nurses and community health workers. This approach will require training, so less specialized healthcare workers can develop the skills needed to identify and make the most appropriate referrals to neurologists and to resume management of stable neurology patients.