In Nigeria, three components of the health care sub-system are the primary, secondary and tertiary levels, governed by the Local government area, State and Federal levels respectively. The secondary health care facility exists to “provide specialised services to patients referred from the primary health care level through out-patient and in-patient services at hospital centers for general, medical, surgical and paediatric patients”(18). Users of secondary health care facilities often live within the same administrative area, which is different from tertiary hospitals where at least half of the clientele reside out of State, seeking highly specialised services and care at such facilities.
Desire to have access to quality health services, competent staff especially being seen by a doctor and access to equipment required for clinical investigation, thereby avoiding subjective diagnosis, are key reasons for seeking higher levels of care in Nigeria(19, 20). However, an assessment of the 25 hospitals in Lagos and Abuja (former and present capital cities of Nigeria) showed that laboratory infrastructure remains weak at secondary health care facilities as only a few of the laboratory staff had received training on pathogen identification, used SOPs or had access to information technology tools, and amplifies the dearth of pathologists and microbiologist. Laboratory diagnostics such as antimicrobial susceptibility testing should guide the management of health conditions caused by bacteria such as sepsis. But, limited access to such services impedes accurate diagnosis and results in poorer health outcomes among patients It also hinders availability of quality microbiologic data to inform empiric treatment guidelines, thus highlighting of strengthening AMR surveillance in Sub-Saharan Africa with evidence-based interventions that use participatory approaches(21, 22).
At the two health facilities where data were abstracted over seven months, less than one in ten of patients in the department studied were found to have suspected sepsis, largely among children 1–5 years from the paediatric outpatient department. The preponderance of sepsis among paediatric patients is similar to other local studies in Nigeria and low-resource settings (23–25). This is likely the result of poor access to vaccines and other prevention and control measures for common preventable infections among children. However, the finding may also be influenced by sepsis case-finding practices at the facilities. Most patients with suspected sepsis did not have their vital signs documented. This may be due to the lack of patient notes and quality data capture, but it is also related to the choice of clinical criteria which doctors and nurses measure, record and use during clinical examinations. Evaluation of vital signs (i.e., body temperature, heart rate, respiratory rate, and blood pressure) is an important triage examination of both children and adults for early recognition, diagnosis, and management of sepsis (26), particularly for patients admitted in the emergency setting (mainly in overloaded and most resource-limited ones). In low-resources settings, the systemic inflammatory response syndrome (SIRS) criteria (i.e., fever or hypothermia, tachycardia, tachypnea, leukocytosis, or leukopenia) and, more recently, the quick Sequential Organ Failure Assessment (qSOFA) criteria (i.e. increased respiratory rate, altered mentation, decreased systolic blood pressure) have often been cited as screening tools to identify patients with suspected sepsis. However, their usefulness in routine practice has been questioned given that the SIRS criteria have demonstrated high sensitivity but low specificity and the qSOFA was validated on patients that already had suspected infection (7). In a review on the best practices of blood cultures in low- and middle-income countries, Ombelet et al suggest a revised set of clinical indications for sampling blood cultures that could be more feasible in such settings including fever or hypothermia and one sign of severity (e.g., hypotension, confusion, increased respiratory rate, suspicion of severe localised infection, or suspicion of other severe infection) (7).
Identification of the focus of infection is also important in sepsis management and optimisation of treatment, especially for cases where a site of infection can be removed or drained, as seen in abdominal infections and soft-tissue abscesses. Although this study found the gastrointestinal and respiratory tracts to be common foci of infection among the suspected sepsis patients as seen in other Nigerian studies, most patients did not have foci of infection recorded, highlighting again potential areas for improvement during clinical examination and data recording (23, 24). In our study, 1 out of 5 suspected sepsis patients tested positive for malaria parasite. Malaria is a common cause of fever with significant morbidity and mortality in Nigeria (29–31) and fever is also a common symptom of sepsis. While people who contract malaria are at risk of developing sepsis and could potentially benefit from antibiotics, especially in malaria-endemic regions and low resource settings like Nigeria (32), the evidence is conflicting. Guidance on diagnostic stewardship and sepsis case finding should also consider malaria diagnostic using the rapid diagnostic test and early treatment while awaiting blood culture result.
Blood culture is considered crucial to isolate microorganisms, guide appropriate antimicrobial therapy, and improve sepsis management. Only 2.7% of all suspected sepsis patients in this study were sent for blood culture, and only 1.9% had an actual blood sample drawn. This is lower than the findings from a previous study in Nigeria where about 12.5% of the patients who met sepsis diagnostic criteria had a blood culture to guide therapy (27). It differs even more significantly from findings in high-resource settings such as in the study by Otto et al. where more than 80% of patients were reported to have had blood sampling done for cultures (33). The low rate of blood culture requests may be attributed to the fact that patients are expected to pay for blood culture diagnostics out-of-pocket which is often a financial burden that cannot be met and may be wariness of clinicians in making this request due to the long turnaround time for blood cultures. According to this study's multivariate analysis, inpatients and those from the urban health care facility were significantly more likely to have a blood culture requested despite the fact that there were more suspected sepsis patients in outpatients compared to in-patients whereas health insurance status was not associated such requests, suggesting that access remains an overall issue for all patients. The fact that there were more suspected sepsis patients in outpatients compared to in-patients, yet inpatients were more likely to have a blood culture request may be attributed to severity of disease. However, Such poor utilisation of blood culture diagnostics has been shown to contribute to delayed patient recovery, missed diagnosis of sepsis resulting in delay in the institution of targeted antibiotics and long hospitalization (34). Diagnostic stewardship improvement strategies should consider advocacy approaches towards health care facilities, particularly secondary level health care facilities; and provision of essential commodities such as blood culture bottles, in order to improve access and availability of quality diagnostics.
Among the limited number of isolates, the most common causative agents of sepsis found in this baseline study were Staphylococcus aureus, Coagulase-negative Staphylococci, Klebsiella spp. and Escherichia coli, a similar distribution to other studies in Nigeria (32, 33). Overall, only one-fifth of blood culture samples in this study yielded growth. Low positive yield may be due to high contamination rates and quality issues along the pre-analytic and analytic pathway, e.g., inadequate asepsis during sample collection, sub-optimal transport and handling of samples, as has been seen in large hospital studies (37). These high rates of contamination along with long result turnaround time or lack of reporting causes mistrust in the diagnostic pathway and results. Accordingly, diagnostic stewardship strategies should include not only laboratory quality improvement efforts but also improved communication mechanisms to build trust between clinicians and laboratory scientists.
Low positive culture yields are also influenced by the use of antibiotics by patients before presenting to facilities (37). In our baseline study, 3 out of 20 suspected sepsis patients reporting to the healthcare facilities were already on antibiotics pre-consult. This could indicate the antibiotic misuse i.e. procuring such prescription-only medicines over-the-counter antibiotics (38). After consultation, the proportion of patients on antibiotics increased to 14 out of 20 patients. In addition, two-thirds of the antibiotics used in our study were in the Watch category. This is in contrast with the target set by the World Health Organization for measuring appropriate, which is that Access antibiotics should constitute 60% of antibiotic consumption by 2023(39, 40). Increased use of Watch antibiotics and broad-spectrum or high-priority agents such as cephalosporins have been reported in Nigeria and other low-resource settings (41–44), although the overuse of Access antibiotics, often first and second choice therapy for common infections, has also been described in such settings (45, 46). Such findings have important local implications for antimicrobial stewardship programs and prioritisation of guidance.
This baseline study is a large-scale study and arguably the first of its kind in Nigeria. The use of standard tools for data collection makes it reliable. However, there are limitations that should be considered. In the first stage, the online survey was completed by the heads of laboratories at 25 out of 65 public secondary health care facilities, so results may have been influenced by non-response bias. In the second stage, although the selected facilities reported regularly using EMR systems, the extraction of data on certain variables such as patients’ clinical characteristics was often incomplete. This was likely due to both lack of measurement during clinical examinations and poor data recording. Laboratory data were extracted from the EMR and manual laboratory registers, but this does not exclude the possibility of missing data and misclassification bias.