Article Text

Download PDFPDF

An analysis of the clinical practice of emergency medicine in public emergency departments in Kenya
  1. Benjamin W Wachira1,
  2. Lee A Wallis2,
  3. Heike Geduld3
  1. 1University of Cape Town, Boston, Bellville, South Africa
  2. 2Division of Emergency Medicine,University of Stellenbosch, Cape Town, South Africa
  3. 3Division of Emergency Medicine, University of Cape Town, Cape Town, South Africa
  1. Correspondence to Dr. Benjamin W. Wachira 21 Salisbury Road, Boston, Bellville 7530 South Africa; benjawambugu{at}


Objectives To describe the case mix, interventions, procedures and management of patients in public emergency departments (ED) in Kenya.

Methods An observational study over 24 h, of patients who presented to 15 public ED during the 3-month period from 1 October to 31 December 2010. The study was conducted across Kenya in two national referral hospitals, five secondary level hospitals and eight primary level hospitals. All patients presenting alive to the ED during the 24-h study period that were seen by a doctor or clinical officer were included in the study. A data collection form was completed by the primary investigator at the time of the initial ED consultation documenting patient demographics, presenting complaints, investigations ordered, procedures done, initial diagnosis and outcome of ED consultation.

Results Data on 1887 patient presentations were described. Adults (≥13 years) accounted for the majority (70%) of patients. Two peak age groups, 0–9 and 20–29 years, accounted for 27% and 25% of patients, respectively. Respiratory and trauma presentations each accounted for 21% of presentations, with a wide spread of other presentations. Over half (58%) of the patients were investigated in the department. 385 patients received immediate treatment in the ED before discharge. Fewer than one in three patients admitted or transferred to specialist units received any therapy in the ED.

Conclusions ED in Kenya provide care to an undifferentiated patient population yet most of the immediate therapy is provided only to patients with minor conditions who are subsequently discharged. Sicker patients have to await transfer to wards or specialist units to start receiving treatment.

  • Case mix
  • emergency care systems
  • emergency department
  • emergency medicine
  • demographics
  • Kenya

Statistics from

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.

In low-income countries, the dominant causes of death are infectious and parasitic diseases (including malaria) and perinatal conditions.1 Kenya has the leading economy in East Africa, but is listed as a low-income country by the World Bank.2 The burden of communicable diseases is high, with malaria as the leading cause of morbidity (30%)3 followed by respiratory diseases (24.5%). HIV/AIDS disproportionately affects the country's mortality and morbidity, with a prevalence rate of 7.4%.3

In middle and high-income countries, the 10 leading causes of death are dominated by non-communicable conditions, which include road traffic accidents.1 Between 2007 and 2050, the world population is expected to increase by 2.5 billion.4 At the same time, the population living in urban areas is projected to gain 3.1 billion, reaching 6.4 billion.4 Africa, in particular, is projected to see its urban population increase by 0.9 billion.4 Population growth is becoming largely an urban phenomenon concentrated in the developing world. At a time when there is much discussion about globalisation, health itself has also become globalised.5 The increasing consumption of tobacco, alcohol and processed foods and changing patterns of living, together with global population ageing, are associated with a rise in the prominence of non-communicable diseases such as cancers, heart disease, diabetes and other conditions linked to obesity.

Already common in industrialised nations (and accounting for 45% of the adult disease burden in low and middle-income countries globally), they now have ominous implications for Africa, which is still dealing with the traditional problems of poverty such as undernutrition and infectious diseases.1 5 The non-communicable disease burden in Kenya is on the increase, with diabetes prevalence at 3.3%, a threefold increase over the past 10 years.3 Mental illnesses and road traffic injuries are also on the increase.3 The emergency workload in Kenya is thus likely to increase.

Emergency medicine is not a specialty in Kenya. Most of the emergency departments (ED) are run by clinical officers who work independently or alongside medical officers to provide healthcare services to largely rural populations. Clinical officers are not doctors but healthcare providers who undergo 3 years of rigorous training following the medical model, and like medical officers, lack specific training in emergency medicine. They are trained in basic sciences, nursing care and have clinical rotations through most inpatient departments. On graduation, they receive a diploma in clinical medicine, surgery and community health, and work as full-time interns for 1 year before getting a licence to practise medicine independently. Internship involves 3-month supervised rotations in the major clinical departments including ED. The Kenya Medical Training College offers post-basic courses to clinical officers, leading to a specialised qualification in several areas, but emergency medicine is not one of them.

There are very few published data on patients seen in ED in Kenya or their clinical management. Furthermore, there is a deficiency in the local medical training on principles of triage and emergency patient management. This study aimed to evaluate the case mix of patients and the current clinical practices in ED across the country, with the hope of informing the development of a sustainable emergency care system appropriate for the local healthcare system.


Study location

The Republic of Kenya, located in eastern Africa, consists of eight provinces: Central, Coast, Eastern, Nairobi, North Eastern, Nyanza, Rift Valley and Western.6 It has a population of over 38 million people, with nearly 68% of Kenya's population living in rural areas.7

There are 222 public primary hospitals, 10 public secondary hospitals and three public tertiary hospitals in Kenya.8 To represent the population distribution, this study was conducted in a public primary hospital in all eight provinces, five public secondary hospitals and two public tertiary hospitals, all randomly selected per province using a random number generator for Excel 2.0.

Data collection

Study data were collected prospectively for one period of 24 h per hospital. Data collection days for each hospital were randomly generated between 1 October and 31 December 2010. Table 1 shows the dates the data were collected in the different hospitals.

Table 1

List of hospitals and data collection dates

All patients presenting alive to the hospitals' ED who were seen by a doctor or clinical officer were included. The principal investigator filled in the information directly onto the data sheets from the patients, and any outstanding information was completed from the patients' records. Data collected on each patient included: age and sex; presenting complaint; investigations ordered; clinical management, including medications and procedures performed; diagnosis (or preliminary diagnosis) and disposition. Most ED in Kenya do not keep records of patients seen in the department. The only records kept are admissions to wards. These admission records were used to cross-reference the admission data in the study.


Patients were excluded if they had incomplete data. This usually related to patients who left the ED after consultation without treatment. Multiple visits by the same patient during the study period were excluded.

Data analysis

Data were entered and analysed using Microsoft Excel 2007 software.


Data were collected for 1887 patient presentations. There were 2011 patient visits for the period, giving a response rate of 93.8%. Tertiary hospitals saw the most patients per day (147 and 152). Secondary hospitals saw an average of 122 patients (87–142) and primary hospitals 122 (13–188) patients per day. Table 2 shows the distribution of patients in the different ED across the country.

Table 2

Number of patients seen per hospital

Patient demographics

There were 1321 (70%) adult patients (≥13 years). There were 947 female patients (50%). The age distribution of the patients is shown in figure 1; 27% of the patients (n=516) were aged 0–9 years and 24% were aged 20–29 years (n=466).

Figure 1

Age distribution of patients.


A total of 1088 patients (58%) had investigations done in the ED. The distribution of investigations performed on patients is shown in figure 2.

Figure 2

Investigations ordered for patients.

A total of 323 patients had x-rays, 66 had ultrasounds, 18 had CT scans and 18 had other imaging tests done. Only nine patients had ECG performed.

One hundred and seven patients had urine tests done, 53 had other specimen tests and 20 had sputum tests. No patient had a lumbar puncture for cerebrospinal fluid analysis performed.

Ten per cent of patients (n=190) had more than one investigation carried out.

Patient presentations/diagnoses

Trauma presentations accounted for 396 (21%) of all presentations. Non-trauma presentations included 393 respiratory (20.8%), 290 abdominal (15.4%), 155 neurological (8.2%), 94 gynaecological (5%), 87 ophthalmological, ear, nose and throat and dental problems (4.6%), 45 sepsis and wound care problems (2.4%) and 20 cardiovascular presentations (1.1%). A further 233 had more than one system involved (12.3%), and 174 with other systems and miscellaneous complaints (9.2%) made up the balance of presentations.

The 10 most common documented diagnoses in adults (accounting for 71% of all adults) and the five most common diagnoses in children (90% of all children) are shown in table 3. Three hundred and seven adults and 134 children had more than one provisional diagnosis.

Table 3

Most common documented diagnoses in adults (2a) and children (2b)

Clinical practice

Only 545 (29%) of patients seen received any intervention in the ED. The five most common treatments offered are shown in table 4.

Table 4

Common treatments in ED in Kenya


Seventy four per cent of the patients seen in the ED (n=1391) were discharged and 19% were admitted to the wards (n=354). Seven per cent of patients (n=127) were referred to specialist clinics or another hospital for management, and eight patients were taken straight to theatre (<1%) for emergency surgery. Seven patients died in the ED (<1%).

Of the 545 patients treated initially in the ED, 385 were discharged (71%), 118 were admitted (22%), 33 were referred (6%), four had emergency surgery (<1%) and five died (<1%).


This is the first description of the profile and outcomes of emergency presentations in Kenyan hospitals. These results will help to inform the development of emergency care systems in the country. Although only 15 hospitals were surveyed, there is no reason to suppose there were different patient profiles at other ED.

Patient profile

The patient age distribution shows two peaks, in the 0–9 and 20–29-year age groups. This is partly a reflection of the country's population distribution (as the 0–9-year age group forms 30% of the country's population).7 Thirty per cent of emergency presentations were children; this is an expected finding, given the fact that children under 5 years of age are seen in the maternal and child health clinics during the day and only present to the ED after hours. Two of the hospitals evaluated had a separate 24 h ED for paediatrics. Unfortunately, no studies have been conducted evaluating the clinical practices in maternal and child health clinics in Kenya, but surveys of local facilities revealed that the quality of care provided to sick children at the first level of health facilities needed improvement.9 10

The 20–29-year age group (which forms 18% of the country's population)7 may be more representative of the group at greatest risk of trauma; 127 (27%) of the 466 patients in this group presented with trauma.


Blood tests were the most common investigations ordered. Anecdotally, it was noted in this study that blood films for malaria parasites were one of the most common blood tests ordered, but only 59% (n=156) of the patients treated for malaria had a positive smear. The rest were treated for malaria without being tested or with negative results. Similar results have been replicated in other studies,11 illustrating a high reliance on patients' clinical features rather than laboratory investigations. This is due to lack of quality control measures in most rural laboratories.11 Cost limitations, as patients have to pay for the services, may play a role.12

Most x-rays ordered (57%, n=175) were to investigate for trauma. Only 23% (n=63) of patients managed for lower respiratory tract infections or pulmonary tuberculosis had x-rays performed, with the majority of patients being managed on a clinical basis.

Focused training for ED staff on the rational use of investigations to support their clinical skills might improve the quality of care and reduce the cost of accessing health care for patients by decreasing inappropriate testing. With reduced workloads, diagnostic services could then focus on quality control measures and improved service provision.

Patient presentations/diagnoses

Trauma, respiratory tract infections and malaria continue to be the leading presentations in ED in Kenya, in keeping with 2004 WHO data.1 HIV/AIDS is the leading cause of morbidity in Kenya.1 Only approximately 1% of the patients in this study (n=17) were noted to have HIV/AIDS at presentation. This may be due to many immunocompromised patients presenting to the ED with other complications, not volunteering their HIV/AIDS status due to the associated stigma.

Malaria was the second most common diagnosis overall (n=266); 61% of these patients (n=162) presented with fever, making it the most common presenting complaint in patients diagnosed with malaria. Interestingly, only 10% of all patients seen (n=185) had their temperature measured in the ED.

Presenting conditions in the ED were varied, both medical and surgical. This demonstrates a need for the attending staff to have comprehensive knowledge and skills required to provide immediate or urgent medical interventions to the undifferentiated population of patients irrespective of their age, gender, or condition.

Clinical practice

All patients with fractures secondary to trauma were managed by the orthopaedic clinical officers. In most institutions, orthopaedic clinical officers are available 24 h a day, but when they were unavailable, patients were asked to come back the following day or were admitted to the wards. Less than a quarter of the patients who presented with fractures received any analgesia or sedation, even for the reduction of the fractures.

Most patients were investigated for malaria and typhoid without consideration of patient presentation. Six per cent of patients with varied presentations (n=110) were treated for malaria without a positive malaria screening test, suggesting an overdiagnosis of the condition. Unfortunately, a positive malaria test meant patients were labelled as such with little consideration of the clinical presentation or the presence of other co-infections. For example, a 20-year-old female patient presenting with confusion and strange behaviour had a positive malaria and typhoid test and was discharged home on treatment for both conditions without any other investigations to rule out other conditions (eg, meningitis).

Unfortunately, we have no means of establishing the bounce-back rate or the outcomes of these patients. The absence of local guidelines developed for the ED indicate a lack of recognition of their value in assisting healthcare providers to stabilise patients who have a life or limb-threatening injury or illness.

Consultation outcomes

Eighteen per cent of the patients seen were admitted or transferred to specialist units or a higher level hospital (n=335) without any immediate therapy in the ED. Some of these were patients admitted or transferred for potentially life-threatening conditions, including sepsis, seizures, severe head injury and organophosphate poisoning.

Anecdotally, this practice is based on the notion that very sick patients or any patient needing admission should not be managed in the ED, but rather should be immediately admitted to the wards so that they can be investigated and treated by the admitting doctor or clinical officer.

Although access to definitive care was accelerated by this model, none of the hospitals had doctors or clinical officers in the wards 24 h a day, and thus very sick patients were left to the care of the ward nurses while awaiting review.


The study period was not long enough to evaluate weekly or seasonal trends in the different hospitals. However, review of epidemiology shows little variation in patient presentation rates or likely patient profiles.13 It was considered unlikely that the outcomes of the study were influenced by the timing of the data collection period.

Assessment of the quality of clinical care, the accuracy of the diagnosis and the need for interventions was not part of the scope of this study. It was not possible to follow-up patients who left the system.


This study highlights a number of important features of ED presentations and patient management in public hospitals across Kenya, which have implications for the development of emergency care systems in the country. Key areas include triage, to reduce the caseload of minor cases, thus allowing adequate resource allocation to very sick patients, and specific emergency medicine training of clinical officers and doctors working in ED on the rational use of investigations and management of acute and urgent illness and injury.


This study would not have been possible without the assistance of the healthcare providers and clerks at all the hospitals studied and Mrs FN Wachira.



  • Funding All costs were borne by the primary investigator.

  • Competing interests None.

  • Ethics approval The study was approved by the University of Cape Town ethics committee (REC 208/2010). Patient confidentiality was maintained by secure storage of data sheets, use of an anonymous coding system and restricted access to the database.

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

Linked Articles

  • Primary survey
    Steve Goodacre