Article Text

Download PDFPDF

Cross-sectional evaluation of emergency care capacity at public hospitals in Zambia
  1. Chancy Chavula1,
  2. Jennifer L Pigoga1,2,
  3. Muhumpu Kafwamfwa3,
  4. Lee A Wallis4
  1. 1 Division of Emergency Medicine, University of Cape Town, Cape Town, South Africa
  2. 2 Emory University School of Public Health, Atlanta, Georgia, USA
  3. 3 Zambian Ministry of Health, Lusaka, Zambia
  4. 4 Division of Emergency Medicine, University of Cape Town, Bellville, South Africa
  1. Correspondence to Jennifer L Pigoga, Division of Emergency Medicine, University of Cape Town, Cape Town 7935, South Africa; jennpigoga{at}gmail.com

Abstract

Objectives The last decade has seen rapid expansion of emergency care systems across Africa, although they remain underdeveloped. In Zambia, the Ministry of Health has taken interest in improving the situation and data are needed to appropriately guide system strengthening efforts. The Emergency Care Assessment Tool (ECAT) provides a context-specific means of measuring capacity of healthcare facilities in low- and middle-income countries. We evaluated Zambian public hospitals using the ECAT to inform resource-effective improvements to the nation’s healthcare system.

Methods The ECAT was administered to the lead clinician in the emergency unit at 23 randomly sampled public hospitals across seven of Zambia’s 10 provinces in March 2016. Data were collected regarding hospitals’ perceived abilities to perform a number of predefined signal functions - life-saving procedures that encompass the need for both skills and resources. Signal functions (36 for intermediate facilities, 51 for advanced) related to six sentinel conditions that represent a large burden of morbidity and mortality from emergencies. We report the proportion of procedures that each level of hospital was capable of, along with barriers to delivery of care.

Results Across all hospitals, most of the level-appropriate emergency care procedures could be performed. Intermediate level (district) hospitals were able to perform 75% (95% CI 73.2 to 76.8) of signal functions for the six conditions. Among advanced level hospitals, provincial hospitals were able to perform 68.6% (67.4% to 69.7%) and central hospitals 96.1% (95% CI 93.5 to 98.7) Main failures in delivery of care were attributed to a lack of healthcare worker training and availability of consumable resources, such as medicines or supplies.

Conclusion Zambian public hospitals have reasonable capacity to care for acutely ill and injured patients; however, there is a need for increased training and improved supply chains.

  • emergency care systems
  • global health
  • emergency departments

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Key messages

What is already known on this subject

  • Emergency medicine is rapidly expanding in limited middle-income countries, but most data on capabilities come from single centre studies.

  • Prior to this study, little was known about the capacity of public hospitals in Zambia to provide adequate emergency care.

What this study adds

  • We used the Emergency Care Assessment Tool (ECAT), to determine the capacity of district, provincial and central hospitals in Zambia to deliver emergency care.

  • Although overall capacity was reasonable, the ECAT identified several common barriers to delivery of emergency care across Zambia, such as lack of emergency care-specific healthcare worker training and a poor supply chain.

Introduction

Robust emergency care systems allow limited resources to be most appropriately targeted towards patients with the best chance of benefitting: those at the greatest risk for morbidity and mortality for which interventions can improve outcomes.1 2 Africans suffer the highest rates of injury and illness, making emergency care services especially important.1 3 As a clinically- and cost-effective means of reducing morbidity and mortality in low-and middle-income countries (LMICs), emergency care systems are quickly expanding across Africa.4–6 Ongoing, rigorous evaluation is critical as these systems develop and mature, to inform decision-making at all levels, from hospital administrators to national policymakers.4 Yet, to date, there have been few assessments of the capacity of African healthcare facilities to provide emergency care.7 8

Zambia is a rapidly-growing lower middle-income country in Southern Africa. The majority of its population lives in rural areas, and below the poverty line.9 There are high burdens of both injury and illness in the country, leading to significant morbidity and mortality.9 10 Most Zambians access healthcare through the public system, which includes three levels of hospital care: district, provincial and central. District hospitals service smaller regions with basic medical, surgical and obstetrical services, and provincial hospitals provide larger populations with slightly more specialised care. Central hospitals are fewest in number, but offer the most specialised services as referral centres.9 Despite there being a total of 99 public hospitals (the distribution of which is described in figure 1), healthcare provision remains inadequate.11

Figure 1

Zambian public hospitals by service level.

The Zambian Ministry of Health (MoH) has begun to prioritise strengthening emergency care services at hospitals across the country. This is largely coming in the form of education. Currently, there are no formal training programmes for emergency medicine specialists, or for emergency nurses and mid-level providers. Recent initiatives in-country, such as the WHO’s Basic Emergency Care course12 and the Emergency First Aid Responder programme,13 are introducing healthcare workers and community members to the concept of emergency care via short courses.

In an effort to understand the current status of emergency care in-country, forming a baseline for larger system improvement, the Zambian MoH requested that the  Emergency Care Assessment Tool (ECAT) be utilised to evaluate facility-based emergency care capacity nationwide.

Methods

We undertook a cross-sectional study of a convenience sample of Zambian public hospitals using the ECAT in March 2016.

Although many tools exist to evaluate emergency care systems, nearly all are designed for well-established systems in high-income countries, and are likely inappropriate for use in LMICs.14 These tools do not take into account differences in delivery of emergency care between high-income countries and LMICs, and were developed in the context of different burdens of disease and injury.15 16 Additionally, most existing tools are not comprehensive: they do not cover the breadth of emergency care, or they assess only one component of care (such as trauma).17–19 The ECAT was developed to fill this void in the African context: it is a standardised, validated and context-specific tool intended to more broadly assess emergency care provisions across Africa,20 21 and one that can be leveraged by countries both before and during development of their emergency care systems.

ECAT is a survey-based tool that uses signal functions to identify a facility’s strengths and weaknesses in delivering emergency care for six sentinel conditions; these were selected via consensus to represent common clinical presentations for which emergency intervention can produce a positive change in patient outcome (online supplementary appendices 1 and 2).22 It evaluates a facility’s ability to perform a varying range of signal functions - life-saving procedures that encompass the need for both skills and resources - based on its designated status as a basic, intermediate or advanced facility.16 The essential list of signal functions for intermediate facilities represents the core of hospital-based emergency care, whereas the essential list for advanced facilities starts to include more skilled and technical interventions. For intermediate facilities (district hospitals), there are 36 signal functions associated with six sentinel conditions, while advanced facilities (both provincial and central) were evaluated based on 51 signal functions across the same six sentinel conditions (table 1). Success is defined as being able to perform a signal function at least 90% of the time.

Supplemental material

Table 1

Total number of signal functions evaluated, based on facility level

Where a facility is unable to perform a signal function at this rate, clinicians are asked to identify at least one of six barriers to delivery: lack of policies; human resources; healthcare worker training; supplies, equipment and medications; infrastructure and no indication. Free text space is also provided to write in any unlisted barriers to delivery. Where appropriate, multiple barriers may be selected for a single signal function.

In order to reflect urban and rural settings across all districts and all tiers of facilities-based care, we aimed to conduct the ECAT at a convenience sample of 25 (approximately 25%) of all public hospitals. Within each tier, we aimed to sample at least 10% of facilities. The ECAT was designed for all hospitals regardless of sector but, since the overwhelming majority of Zambians access healthcare through the public system,9 public hospitals were prioritised for this study. Basic facilities - health posts and health centres - were not included in this study.

All randomly selected sites were contacted by Ministry of Health officials via telephone and letter regarding the study and visit dates. Trained researchers followed a standardised procedure for collecting data. The lead physician working in the emergency unit (EU) was identified and verbally informed of the study, after which written informed consent was obtained. Participants remained anonymous. The researcher provided written and verbal survey instructions to the clinician, along with a paper copy of the assessment tool. The ECAT was then completed by the clinician; the researcher remained available to clarify as needed.

Survey completion time typically ranges from 45 min to 2 hours. Data were entered into encrypted Microsoft Excel (Microsoft, Redmond, Washington) spreadsheets, after which basic and inferential statistics were generated using SAS V.9.4 software (SAS, Cary, North Carolina).

Results

A total of 23 public hospitals were surveyed: seven district, 12 provincial and four central. Due to weather conditions during the study period, two provinces (North-Western and Luapula) were not easily accessible by road, and the two selected hospitals were non-responsive to email and telephone contact. As we could not reach them to discuss participation they were excluded from the study. Therefore, only 23 of the 25 randomly-selected facilities were visited during the study period. In line with study goals to assess at least 10% of each facility type, the ECAT was conducted at approximately 10% of district hospitals, 57% of provincial hospitals and 67% of central hospitals across seven of Zambia’s 10 provinces. Given that provincial and central hospitals are located in more urban, centrally-located areas of the country, additional facilities in these categories beyond the 10% goal were sampled out of convenience.

Signal function performance varied across facility-level designations and sentinel conditions (figure 2). The intermediate ECAT assessment is comprised of 36 signal functions across the six sentinel conditions. Central facilities performed better than provincial and district in all condition categories; provincial and district had similar performance.

Figure 2

Mean number of intermediate and advanced signal function performed across facility levels.* *The ECAT does not evaluate any additional skills for the advanced care of severe pain and maternal health. ECAT, Emergency Care Assessment Tool.

The ECAT also lists 15 additional advanced signal functions for four of the six sentinel conditions; there are none designated for severe pain and maternal health. These advanced signal functions could be performed far more consistently at central (44%) versus provincial (83%) hospitals (they were not performed for the district hospitals, which were deemed intermediate level).

As shown in table 2, signal functions associated with maternal health generally had the highest success rate across all facilities, while those associated with trauma had the lowest. District hospitals, which were surveyed using the intermediate-level tool, had a 75.0% success rate across all signal functions. Provincial and central hospitals both utilised the advanced-level tool, which evaluated more signal functions. Central hospitals cited higher signal function success than provincial hospitals: 96.1% at central hospitals versus 68.6% at provincial hospitals.

Table 2

Mean percentage of signal functions that can be completed successfully, by facility level

Signal function performance varied depending on the condition: At district hospitals, trauma was lowest (57.1%) while altered mental status, severe pain and maternal health were highest (approximately 85%). Provincial hospitals performed lowest for respiratory failure (46.9%) and highest for maternal health (91.7%); central hospitals ranged from trauma (89.6%) to respiratory failure, shock and maternal health (100%). Central hospitals outperformed district and provincial hospitals in every sentinel condition category, the most marked of which was trauma.

At the district level, escharotomy was the trauma signal function with the lowest success rate (8.3%), followed by cooling care (for burns) and cervical spine immobilisation (33.3%). At provincial hospitals, autotransfusion from chest tubes and access to neurosurgical services (0% and 16.7%, respectively) had the lowest rates of success within the trauma functions. Central hospitals had fewer notable trauma-related failures: 25% could perform autotransfusion from chest tubes and 75% could perform thoracotomies and had access to neurosurgical services at this level of hospital.

Data on barriers to delivery were collected on five of six sentinel conditions: no data were captured for maternal health due to a printing error in the ECAT. Indications were given for all failures at district and central hospitals, and 98% of failures at provincial hospitals. Healthcare worker training was the most frequent rationale provided for being unable to perform four of the five signal functions assessed for failure. Across all conditions and facilities, healthcare worker training was the most common barriers to delivery (cited in 66% of failures), followed by availability of supplies, equipment and medications (cited in 49% of failures) (table 3).

Table 3

Percentage of performance failures attributed to each category of failure across all facilities

Frequencies of barriers to delivery at district hospitals ranged from 1% (infrastructure) to 45% (healthcare worker training). At provincial hospitals, healthcare worker training was again the most common barrier to delivery (associated with 63% of failures), while policy was the least common (3%). Barriers at central hospitals were noted in 0% (policy) to 53% (availability of supplies, equipment and medications) of signal function failures (table 4). Advanced signal functions, which were evaluated for only at provincial and central hospitals, were more commonly completed at central (94.4%) versus provincial (56.0%) facilities. Provincial hospitals struggled most with the advance functions associated with the care of respiratory failure (80.6% failure), followed by those associated with caring for shock (25.0% failure). As with intermediate-level signal functions, the most commonly noted barriers to care were healthcare worker training (associated with 71% of failures), and availability of supplies, equipment and medications (61%).

Table 4

Percentage of performance failures attributed to each category of failure, by facility level

Discussion

Zambian hospitals report being able to perform most of the skills necessary to manage life-threatening emergencies. Although two hospitals in the 25% sample were not able to participate, the study is still nationally reflective: seven of Zambia’s nine rural provinces, and its one urban province, were included. The ECAT was simple to implement in the Zambian context and generated uncomplicated results that could be rapidly translated into action items for the Ministry of Health. These exact results may not directly translate to other sub-Saharan African nations, but these data, and the study protocol itself, contribute substantial guidance for the sustainable development of emergency care systems across most LMICs.

Based on previously-established emergency care provision expectations across facility levels16; central hospitals have the greatest capacity for providing emergency care, as was anticipated. Performance in maternal health was high across all facility levels; this may reflect significant investment in maternal health training in Zambia23 that has likely translated practical skills for handling maternal and obstetrical emergencies to most clinicians. The lowest performance across all facilities was associated with trauma care: while central hospitals had high success rates, district and provincial hospitals could perform less than 60% of expected trauma-related signal functions successfully. Trauma care was particularly hindered by the inability to perform more complex procedures, such as thoracotomy, as well as a lack of access to neurosurgical and orthopaedical services. Identified barriers to delivery, particularly for trauma care, suggest that clinicians are allowed to complete the procedures, but are lacking the perceived ability to perform them. Placing more specialised doctors into these facilities is unlikely to ever be realistic but training the existing clinicians in trauma-related procedures appropriate to their facility level (ie, intermediate or advanced) could allow for an improvement in trauma care.

Effective and efficient referral systems have a setup wherein capacity increases as a patient scales the ladder of care. In the context of signal functions, this would mean that more signal functions would be reliably performed as a patient moved from a primary hospital upwards. Furthermore, in order for a patient’s movement from a district to provincial hospital before going to the tertiary central hospital to be worthwhile (as opposed to going directly from district to central, skipping provincial), the mid-level provincial facility must provide more advanced care.

We found that that provincial facilities performed less successfully than central but similar to district hospitals across all intermediate-level signal functions (figure 2). For advanced signal functions (which were expected of only provincial and central hospitals, and exist for four of the six sentinel conditions evaluated), provincial hospitals performed markedly lower than central hospitals; these results are similar to a recent study in Tanzania that found lower-level facilities were less capable of stabilising acute trauma patients prior to referring them to a higher-level facility.24 Given that mid-level hospitals in Zambia are currently unlikely to contribute significantly to a patient’s emergency care, efforts to strengthen the referral chain might be best targeted towards provincial hospitals.

Nearly all failures of signal functions were attributed to at least one barrier; this indicates that clinicians are very aware of what prevents them from delivering appropriate care. The two most notable barriers to delivery – healthcare worker training and availability of consumable resources – align with those found to impede care delivery in other African nations.20 21

This study’s findings are supportive of anecdotal information on the current provision of emergency care training for healthcare workers in Zambia. Steps are currently being taken to introduce the concepts of emergency care within healthcare, including short courses for practising providers as well as nursing and medical students. These efforts are an improvement; however, they cannot serve as a standalone long-term and must rather act as the foundation for more formalised training programmes.

One particular system problem that the ECAT uncovered was a potential weakness in the referral system. At present, high-acuity patients are typically referred from district to provincial hospitals, who may in turn refer them to central hospitals as needed. However, while the data suggest that provincial hospitals perform similarly to district hospitals with regard to intermediate-level procedures (representing the core of emergency care), they lack the capacity to perform more skilled and technical interventions that are expected of them as advanced facilities. Emergencies are time-sensitive, and the additional stop at a provincial hospital that is capable of minimally more than a nearby district hospital might adversely affect patient outcomes. Ultimately, provincial hospitals will need to be up-trained to meet the standards set forth for this level facility and improve signal function performance. But, in the interim, referral systems should be revised such that acute patients are only transferred to facilities that are actually capable of providing a higher level of care for these acute conditions.

In addition to training, our assessment identifies a second gap in emergency care delivery: the deficiency of equipment, supplies and medications. Because the ECAT does not gather specific details regarding what is unavailable, we cannot discern exactly what items are missing in EUs. The lack of consumables was noted across all sentinel conditions and facility levels, indicating that it is likely an overall supply chain issue. On the other hand, facility-based infrastructure, such as electricity, imaging machines and laboratory facilities, was readily available. Previous studies have found that both supply chain and cost can prevent necessary resources from being available in African EUs6 25; further investigation will be required to understand what is preventing consumables from being available at the units we studied.

Other barriers to delivery, such as existing policies and processes, were assessed as having minimal impact on performing signal functions, in keeping with other ECAT assessments.20 21 Policy limitations are harder to identify, especially in contexts where even the most experienced clinicians have minimal knowledge of, or exposure to, higher-functioning emergency care systems. Given the numerous potential causes of this finding, it warrants future investigation.

Limitations

Some of the observed variance in hospital performance might be due to sample size, given that only one person completed the ECAT per site. Increasing the number of clinicians surveyed could improve confidence but, at present, it is challenging to remove multiple providers from already-strained EUs. As such, assessments are based on limited, self-reported knowledge of signal function capacity. No specific physicians were chosen by the Ministry of Health to participate; choosing the most experienced clinician at each site should help ensure representative results, but potential bias in responses mean that the results are likely a best-case scenario: experienced clinicians might overstate capacity in comparison to their more junior counterparts, but we feel their knowledge of the department at-large outweighs this limitation. While validity has already been established, repeated sampling of multiple personnel to establish inter-rater and intra-rater reliability will need to be prioritised in future studies.

Another, perhaps lesser limitation, to our study is that data on barriers to delivery for maternal health were not captured. Maternal mortality remains a major problem in Zambia and probably reflects a shortage of skilled birth attendants.26 Unfortunately, we are unable to comment on how barriers to emergency care are impacting on this.

This is the largest nationwide assessment of facility-based emergency care capacity in Africa to date and its results, which identify critical barriers to emergency care delivery, have provided useful guidance for the MoH and other stakeholders. Initial investments in emergency care by the Ministry are evident in the infrastructure of facilities and existing policies; however, long-term investments in the healthcare workforce and supply chain need to be prioritised. In its initial response to this study, the Ministry is embarking on a training programme for district and provincial hospitals and development of norms and standards for equipment and infrastructure across all levels of facilities. Regular evaluations will help to continue to improve emergency care in Zambia.

References

Footnotes

  • Contributors All authors contributed to the design of this study. CC and JP collected and analysed data. All authors contributed to the writing of the manuscript and have approved of the final version.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Ethics approval University of Cape Town and University of Zambia Health Research Ethics Committees.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Additional data can be made available by contacting the corresponding author.

  • Patient consent for publication Not required.