A baseline review of the ability of hospitals in Kenya to provide emergency and critical care services for COVID-19 patients

Introduction
As the Coronavirus Disease 2019 (COVID-19) cases in Kenya begin to rise, the number of severe and critical COVID-19 patients has the potential to quickly overload the local healthcare system beyond its capacity to treat people.

Objective
The purpose of this study was to gather information about the ability of hospitals in Kenya to provide emergency and critical care services and to identify priority actions for use by policymakers and other stakeholders as a roadmap toward strengthening the COVID-19 response in the country.

Methods
This was a comprehensive review of the published and grey literature on emergency and critical care services in Kenya published in the last three years through April 2020. Screening of articles was conducted independently by the authors and the final decision for inclusion was made collaboratively. A total of 15 papers and documents were included in the review.

Key recommendations.

There is an urgent need to strengthen prehospital emergency care in Kenya by establishing a single toll-free ambulance access number and an integrated public Emergency Medical Services (EMS) system to respond to severe and critical COVID-19 patients in the community and other emergency cases. Functional 24-h emergency departments (EDs) need to be established in all the level 4, 5 and 6 hospitals in the country to ensure these patients receive immediate lifesaving emergency care when they arrive at the hospitals. The EDs should be equipped with pulse oximeters and functioning oxygen systems and have the necessary resources and skills to perform endotracheal intubation to manage COVID-19-induced respiratory distress and hypoxia. Additional intensive care unit (ICU) beds and ventilators are also needed to ensure continuity of care for the critically ill patients seen in the ED. Appropriate practical interventions should be instituted to limit the spread of COVID-19 to healthcare personnel and other patients within the healthcare system. Further research with individual facility levels of assessment around infrastructure and service provision is necessary to more narrowly define areas with significant shortfalls in emergency and critical care services as the number of COVID-19 cases in the country increase.

Diagnostic assistance to improve acute burn referral and triage : assessment of routine clinical tools at specialised burn centres and potential for digital health development at point of care

Background: Inappropriate referral of patients for specialised care leads to overburdened health systems and improper treatment of patients who are denied transfer due to a scarcity of resources. Burn injuries are a global health problem where specialised care is particularly important for severe cases while minor burns can be treated at point of care. Whether several solutions, existing or in development, could be used to improve the diagnosis, referral and triage of acute burns at admission to specialised burn centres remains to be evaluated.

Aim: The overarching aim of this thesis is to determine the potential of diagnostic support tools for referral and triage of acute burns injuries. More specifically, sub-aims include the assessment of routine and digital health tools utilised in South Africa and Sweden: referral criteria, mortality prediction scores, image-based remote consultation and automated diagnosis.

Methods: Studies I and II were two retrospective studies of patients admitted to the paediatric (I) and the adult (II) specialised burn centres of the Western Cape province in South Africa. Study I examined adherence to referral criteria at admission of 1165 patients. Logistic regression was performed to assess the associations between adherence to the referral criteria and patient management at the centre. Study II assessed mortality prediction at admission of 372 patients. Logistic regression was performed to evaluate associations between patient, injury and admission-related characteristics with mortality. The performance of an existing mortality prediction model (the ABSI score) was measured. Study III and IV were related to two image-based digital-health tools for remote diagnosis. In Study III, 26 burns experts provided a diagnosis in terms of burn size and depth for 51 images of acute burn cases using their smartphone or tablet. Diagnostic accuracy was measured with intraclass correlation coefficient. In Study IV, two deep-learning algorithms were developed using 1105 annotated acute burn images of cases collected in South Africa and Sweden. The first algorithm identifies a burn area from healthy skin, and the second classifies burn depth. Differences in performances by patient Fitzpatrick skin types were also measured.

Results: Study I revealed a 93.4% adherence to the referral criteria at admission. Children older than two years (not fulfilling the age criterion) as well as those fulfilling the severity criterion were more likely to undergo surgery or stay longer than seven days at the centre. At the adult burn centre (Study II), mortality affected one in five patients and was associated with gender, burn size, and referral status after adjustments for all other variables. The ABSI score was a good estimate of mortality prediction. In Study III experts were able to accurately diagnose burn size, and to a lesser extent depth, using handheld devices. A wound identifier and a depth classifier algorithm could be developed with assessments of relatively high accuracy (Study IV). Differences were observed in performances by skin types of the patients.

Conclusions: Altogether the findings inform on the use in clinical practice of four different tools that could improve the accuracy of the diagnosis, referral and triage of patients with acute burns. This would reduce inequities in access to care by improving access for both paediatric and adult patient populations in settings that are resource scarce, geographically distant or under high clinical pressure.

Publicly funded interfacility ambulance transfers for surgical and obstetrical conditions: A cross sectional analysis in an urban middle-income country setting

Introduction
Interfacility transfers may reflect a time delay of definitive surgical care, but few studies have examined the prevalence of interfacility transfers in the urban low- and middle-income (LMIC) setting. The aim of this study was to determine the number of interfacility transfers required for surgical and obstetric conditions in an urban MIC setting to better understand access to definitive surgical care among LMIC patients.

Methods
A retrospective analysis of public interfacility transfer records was conducted from April 2015 to April 2016 in Cali, Colombia. Data were obtained from the single municipal ambulance agency providing publicly funded ambulance transfers in the city. Interfacility transfers were defined as any patient transfer between two healthcare facilities. We identified the number of transfers for patients with surgical conditions and categorized transfers based on patient ICD-9-CM codes. We compared surgical transfers from public vs. private healthcare facilities by condition type (surgical, obstetric, nonsurgical), transferring physician specialty, and transfer acuity (code blue, emergent, urgent and nonurgent) using logistic regression.

Results
31,659 patient transports occurred over the 13-month study period. 22250 (70.2%) of all transfers were interfacility transfers and 7777 (35%) of transfers were for patients with surgical conditions with an additional 2,244 (10.3%) for obstetric conditions. 49% (8660/17675) of interfacility transfers from public hospitals were for surgical and obstetric conditions vs 32% (1466/4580) for private facilities (P<0.001). The most common surgical conditions requiring interfacility transfer were fractures (1,227, 5.4%), appendicitis (913, 4.1%), wounds (871, 3.9%), abdominal pain (818, 3.6%), trauma (652, 2.9%), and acute abdomen (271, 1.2%).

Conclusion
Surgical and obstetric conditions account for nearly half of all urban interfacility ambulance transfers. The most common reasons for transfer are basic surgical conditions with public healthcare facilities transferring a greater proportion of patient with surgical conditions than private facilities. Timely access to an initial healthcare facility may not be a reliable surrogate of definitive surgical care given the substantial need for interfacility transfers.

Outcomes of paediatric patients ventilated in a high-care area outside an intensive care unit

Background. Limited availability of paediatric intensive care beds in the public sector is a major challenge in South Africa. It often results in patients being ventilated in a high-care area (HCA) outside an intensive care setting. The outcomes of paediatric patients ventilated outside a paediatric intensive care unit (ICU) are not well documented.

Objectives. To describe characteristics and outcomes of patients ventilated in a paediatric HCA.

Methods. A retrospective chart review of children (0 – 16 years) requiring mechanical ventilation in the HCA at Chris Hani Baragwanath Academic Hospital, Johannesburg, between 1 February and 31 October 2015 was performed.

Results. A total of 214 patients required mechanical ventilation during the study period. Fifty-four percent were male and 91.1% were HIV-negative. The most common diagnoses were acute lower respiratory tract infections (59.3% of the post-neonatal group, 28.8% of the neonatal group) and sepsis (6.8% of the post-neonatal group, 28.8% of the neonatal group). The ultimate rate of acceptance to an ICU was 69.0%. Only 41.6% of cases referred to an ICU were initially accepted, with limited bed availability being the main reason for refusal. Patients with respiratory illnesses were more likely and those with neurological illness less likely to be accepted to an ICU. Patients with low-risk diagnoses were more likely to be accepted than those with very high-risk diagnoses. The overall mortality rate was 32.2%, with 52.2% of these deaths occurring in the HCA. Patients aged 1 – 5 years had the highest mortality rate (48.0%). Lower respiratory tract infections (36.8%) and sepsis (20.6%) were the main causes of death. The mortality rate of suitable ICU candidates in the HCA was higher than that in an ICU (33.3% v. 24.3%). The standardised mortality ratio (SMR), as predicted by the Paediatric Index of Mortality 3 score, for all patients who died in the HCA was 3.3, while the SMR for patients who died in an ICU was 1.3. The odds ratio for mortality of suitable candidates ventilated in the HCA v. patients who were ventilated in an ICU was 1.80 (95% confidence interval 1.39 – 6.03).

Conclusions. Although a reasonable number of paediatric patients ventilated in an HCA survive, survival is lower than in those ventilated in an ICU. However, offering life-supporting therapies in an HCA may offer benefit where ICU care is unavailable. Emphasis needs to be placed on improving access to ICU care as well as optimising the use of available resources.

Evaluation of a digital triage platform in Uganda: A quality improvement initiative to reduce the time to antibiotic administration

Background
Sepsis is the leading cause of death in children under five in low- and middle-income countries. The rapid identification of the sickest children and timely antibiotic administration may improve outcomes. We developed and implemented a digital triage platform to rapidly identify critically ill children to facilitate timely intravenous antibiotic administration.

Objective
This quality improvement initiative sought to reduce the time to antibiotic administration at a dedicated children’s hospital outpatient department in Mbarara, Uganda.

Intervention and study design
The digital platform consisted of a mobile application that collects clinical signs, symptoms, and vital signs to prioritize children through a combination of emergency triggers and predictive risk algorithms. A computer-based dashboard enabled the prioritization of children by displaying an overview of all children and their triage categories. We evaluated the impact of the digital triage platform over an 11-week pre-implementation phase and an 11-week post-implementation phase. The time from the end of triage to antibiotic administration was compared to evaluate the quality improvement initiative.

Results
There was a difference of -11 minutes (95% CI, -16.0 to -6.0; p < 0.001; Mann-Whitney U test) in time to antibiotics, from 51 minutes (IQR, 27.0–94.0) pre-implementation to 44 minutes (IQR, 19.0–74.0) post-implementation. Children prioritized as emergency received the greatest time benefit (-34 minutes; 95% CI, -9.0 to -58.0; p < 0.001; Mann-Whitney U test). The proportion of children who waited more than an hour until antibiotics decreased by 21.4% (p = 0.007). Conclusion A data-driven patient prioritization and continuous feedback for healthcare workers enabled by a digital triage platform led to expedited antibiotic therapy for critically ill children with sepsis. This platform may have a more significant impact in facilities without existing triage processes and prioritization of treatments, as is commonly encountered in low resource settings.