Inability of doctors to correctly tell apart common illnesses ‘worrying’


Lack of ability and knowledge among health providers to accurately diagnose and treat common illnesses is partly to blame for the growing cases of serious injury and deaths of patients caused by negligence.

Four in five doctors, clinical officers and nurses cannot accurately diagnose all four common conditions, including pneumonia and severe dehydration due to diarrhoea in children, and tuberculosis of the lungs and type 2 diabetes in adults, reveals the Kenya Health Service Delivery Indicator Survey (SDI) 2018 that was released recently. The four illnesses are the tracer conditions in the study that aims to determine the quality of services in basic healthcare.

A higher proportion of doctors can correctly diagnose all four conditions, about a third, followed by clinical officers (about a quarter) and nurses (12 per cent).

Less than half (43 percent) of the providers can correctly diagnose three of the four conditions, more than a quarter are able to accurately detect two cases and 10 percent know how to identify just one illness, according to the survey that was done in 3,094 health facilities across Kenya. About one percent cannot correctly diagnose even one condition.

The rate of correct diagnosis of the tracer conditions varies across different cadres. A higher proportion of doctors can correctly diagnose all four conditions, about a third, followed by clinical officers (about a quarter) and nurses (12 per cent).

The worrying revelation on competency comes against a backdrop of growing complaints over misconduct by health providers. There were at least 115 cases of major medical misconduct, mostly involving doctors (two-thirds), that resulted in the death or injury of patients published by the media in the last 20 months, according to Do No Harm, a NationNewsplex database on medical malpractice compiled from mainstream media stories. Patients died in four out of five of the incidents. In the first eight months of 2019, the number of patients or their family members who complained to the media that they were harmed by the treatment they received from medical facilities was 100, eight times the number in the same period last year (13).

However, data from the Kenya Medical Practitioners and Dentists Board (KMPDB), a statutory authority established to regulate the practice of medicine and dentistry, suggests that many cases of misconduct by doctors go unreported to the regulatory institution. From 1997 to March last year, 985 cases were lodged with the board. Of these, 70 cases were pending and 915 had been determined. About 97 per cent of them were resolved at the preliminary inquiry committee, and two per cent each at the national tribunal and the national professional conduct committee. Over the two decades, only one doctor has been deregistered by the medical board. Four doctors had their licences cancelled while six had theirs suspended for a period of six to 12 months for misconduct over the same period. But a majority walked away with a slap on the wrist, with 16 being sent for supervised training for a period of six to 12 months while 106 were ordered to go through continuous professional development.

Correct treatment

Figures from the World Bank and the Ministry of Health study that was done from March-July last year show that the incompetence of health providers results in misdiagnosis of about a quarter of cases involving the four conditions handled by doctors and clinical officers, and 40 per cent that are managed by nurses.

Diagnostic accuracy rate varies across case conditions, ranging from 97 percent accuracy for pulmonary tuberculosis to 32 percent for severe dehydration. The accuracy rate of pneumonia diagnosis is 82 percent and diabetes (61 percent).

Higher-level facilities can correctly diagnose more cases (hospitals, 76 per cent) than lower-level ones (health centres, 68 per cent, and dispensaries, 64 per cent). Likewise, accuracy is higher for urban (70 percent) than rural (66 percent) facilities and it varies widely across counties.

Providers in Vihiga County can only correctly diagnose about half of the tracer conditions, the worst rate, while their counterparts in Wajir can accurately diagnose nearly 90 percent of the cases, the best proportion in the country.

A correct diagnosis is, however, not a guarantee for providing the correct treatment. There are noticeably large inconsistencies between diagnosis and treatment across the board, and this reveals a critical disconnect in provider knowledge and follow-up, finds the study. For instance, while 97 per cent of health providers were accurate in their TB diagnosis, just 74 per cent got the accurate treatment. This means that almost a quarter of lung TB cases that were correctly diagnosed, were not accurately treated. The results of the other conditions equally show a knowledge gap in clinical diagnosis and patient management.

Interestingly, among cases of severe dehydration and Type 2 diabetes, more providers offered correct treatment even though they had lower diagnostic accuracy.

Coexisting conditions

Additionally, there is a knowledge gap on treatment for coexisting conditions, especially among clinicians. For example, the simulation presenting pneumonia also presented the “patient” with a high temperature. While 79 percent of clinicians treated pneumonia only, 72 percent of doctors and nurses got the full treatment of pneumonia and fever, despite paracetamol (for managing fever) being available in a majority of facilities (91 percent).

Experts say that the challenge of partial treatment has short and long-term effects. For instance, for the case of fever, if the child is not tepid-sponged, their fever could reach higher levels and the child could easily have a convulsion, which may cause serious effects, including aspiration of food or fluids and biting of the tongue. If the convulsion is not well understood, it could also lead to further mismanagement of the condition.

The study recommends that training needs to be better focused to prepare health workers to accurately diagnose and treat the main causes of illness and refer complicated cases up to higher levels of care.

Competence aside, according to the latest SDI, the outpatient caseload was high, with the average health worker seeing on average 13 patients per day, an increase from nine in 2012. Kenya’s caseload was only lower than Mozambique’s (17 patients per day) among nine select African countries including Tanzania (seven), Madagascar (five) and Niger (10).

The caseload is usually of concern because a shortage of health workers may cause it to rise and potentially compromise service quality. Large health facilities (above 20 staff) have very low caseload levels, with fewer than seven patients per provider per day. Small-sized facilities (one to two staff), comprising mostly of dispensaries and clinics, had the highest caseload (15 outpatients per day), implying how health workers could be reallocated and be better used in primary health facilities with higher caseloads. Public facilities had twice the daily caseload (18 patients per provider per day) of private ones (eight).

Rural facilities on average had four more patients a day than urban facilities. While lower-level facilities had higher caseloads in the public sector, the situation was the opposite in private institutions.
The caseload was no doubt worsened by the high absenteeism. According to the survey, the absence rate was more than half (53 percent) during unannounced visits, with the public sector absenteeism at 57 percent compared to 48 percent in the private sector. Doctors had the highest absenteeism rate of about two-thirds (61 percent) followed by nurses (55 percent) and clinical officers (50 percent).

Absences unauthorised

Although some of the absentees had permission, from a patient’s perspective, these providers are not available to deliver services at the health facility – whether authorised or not. Overall, less than half of the absences were authorised (46 percent) while 11 percent were unauthorised. Nine percent of health workers were on official mission, seven percent were on training and five percent were on medical leave.

This finding reveals that there is a clear need for better organisation and management of skilled human resources to improve the availability of staff to offer services, especially as it relates to authorised absenteeism.

However, according to data from the KMPDB from 1997-2018, there were only a handful of complaints submitted related to absconding duty. Among the cases captured by Do No Harm is that of Sharon Nyambura, who died in June 2018 at age 22, after being rushed to a private hospital in Kiambu to give birth. According to her family, she started bleeding and fainted as staff were looking for doctors to attend to her.

In another case reported last month, 26-year-old Portia Lavine and her unborn child passed away just hours apart at Kenyatta National Hospital, where she had arrived by ambulance after being referred from Pumwani Hospital. She waited for six hours to be attended to by a doctor, who only showed up when she started vomiting blood.

Kenya performed below average on absenteeism and equipment, with the country’s absence rate being quadruple that of Tanzania and double that of Mozambique and Madagascar.

The health providers also compromised the quality of healthcare by routinely disregarding medical guidelines. Less than half of the providers (44 percent) adhered to clinical guidelines in the management of the four tracer conditions. Urban providers were more adherent to the guidelines (46 percent) than rural ones (42 percent). Doctors abided to more of the clinical guidelines (52 percent) followed by clinical officers (48 percent) and nurses (38 percent). Adherence to clinical guidelines was lowest in dispensaries and clinics (41 percent) followed by health centres (44 percent) and hospitals (50 percent). For the most part, clinical guidelines are not followed in primary-care health facilities, which is usually the first point of entry for most beneficiaries.


You're all set to enjoy unlimited Prime content.