Lumbar puncture refusal in sub-Saharan Africa: A call for further understanding and intervention

In 1891, Winter¹ described the first 4 cases of tuberculous meningitis (TBM), in which “paracenteses of the theca vertebralis was performed to relieve cerebrospinal fluid (CSF) fluid pressure.” Since this original description of the lumbar puncture (LP) procedure, neurologists worldwide have relied on LPs for both diagnostic and therapeutic purposes. In resource-limited settings, LPs are often the only neurologic test available to aid the clinician in neurologic diagnosis. In sub-Saharan Africa, a large number of patients present to hospitals with acute neurologic symptoms, including those who are HIV-infected and have opportunistic infections such as cryptococcal meningitis and TBM. In these clinical scenarios, LPs are an essential point-of-care diagnostic and therapeutic procedure.² The benefits of the LP as a diagnostic tool are well-known, but it is important to emphasize that therapeutic LPs are a low-cost measure to monitor and treat intracranial pressure (ICP) due to nonobstructive hydrocephalus in regions of the world where more sophisticated testing and treatment are unavailable due to limitations of medical equipment, medication supplies, and clinical personnel, including specialized neurologists and neurosurgeons.

Procedural training as well as LP equipment and laboratory capacity to run basic CSF studies is required for safe and optimal use of the procedure. Although these tools are often available in major medical centers in sub-Saharan Africa, there have been reports that LPs are being deferred among a substantial number of patients presenting with neurologic symptoms that warrant CSF evaluation. Immediately prior to the Cryptococcal Optimal Antiretroviral Therapy Timing (COAT) Trial,³ refusal of a diagnostic LP was 25% at Mulago Hospital in Kampala, Uganda (personal communication, Ann Vogt and David Boulware, 2014). In a study of inpatient pediatric care in 14 district hospitals in Kenya, no LPs were performed on 46 patients presenting with meningitic symptoms and signs, despite 13 hospital laboratories having capacity for CSF microscopy.⁴ In another study of new-onset seizures in HIV-infected patients in Lusaka, Zambia, 12% of patients refused diagnostic LP despite efforts to educate patients on the procedure's possible benefits (personal communication, Omar Siddiqi, 2014). These high refusal rates raise several questions, including whether the reasons for LP refusal are region-, sex-, or disease-specific, and whether refusals are driven by patients' perceptions or due to practitioners' discomfort with the procedure.

An informal 8-question survey was sent to 10 clinicians practicing in urban medical centers in Nigeria, Malawi, and Zambia regarding the issue of LP deferral. Eight clinicians (80%) responded, with 4 (50%) respondents stating that LP refusal was a major barrier to clinical care in their hospital setting. All respondents practicing in Zambia and 1 practicing in Northern Nigeria believed that this was a major clinical issue. Two respondents in Zambia reported that this was an issue faced daily in clinical practice. Four practitioners cited that LP refusal was related to patients' and relatives' preexisting belief that the LP procedure will cause death. A common perception is the association of LP with death, due to the high mortality associated with meningitis, with misattribution of the LP procedure causing death due to the underlying condition. This association is believed to be strongest in the setting of HIV-infected patients who require therapeutic LPs for cryptococcal meningitis, the most common cause of meningitis in Eastern and Southern Africa. Another reason cited by 3 practitioners for LP deferral was practitioners' discomfort with performance of the procedure and in particular, assessment of the risk for herniation without access to neuroimaging. Four clinicians cited fear of herniation as a reason for not performing LPs. Evidence suggests that the risk of herniation due to the LP procedure is rare. In a retrospective study in the pre-CT era of 129 patients with varying pathology, 70 with papilledema and 59 without papilledema but with increased CSF opening pressure on LP, 1 patient with papilledema died presumably due to the LP.⁵ Among 832 persons with meningitis in Uganda and South Africa, none died of herniation due to LPs.⁶ In countries where neuroimaging is easily accessible, there remains no consensus on the indications of CT before LP, or on the effectiveness of CT to detect patients in whom clinical deterioration may occur with a LP.^7,8 Clinician discomfort may also be related to lack of training in ophthalmoscopy skills and the neurologic examination or lack of proper medical equipment.

Patients' refusal of LPs may also be due to lack of explanation of the indication as well as the risks and benefits of the procedure. The root cause of misperceptions has not been explored, and the lack of routine formal informed consent is a missed opportunity to correct misperceptions. In all locations surveyed, no formal written consenting process is performed, outside of research studies. In addition, lack of confidence in accurate laboratory results lessens clinicians' enthusiasm for performing LPs if the results are untrustworthy. Depending on the country and location, the quality of laboratory facilities may be highly variable, with a lack of standard laboratory operating procedures, no internal or external quality assurance testing, or varying skill of supervisors and technical personnel.⁹ Additionally, clinicians with little or no ancillary support in busy hospital settings may believe that the time needed to perform an LP may outweigh its potential diagnostic and therapeutic benefits. However, all clinicians surveyed readily recognized the consequences of not performing LPs, including delay in diagnosis and lack of proper treatment, resulting in increased morbidity and mortality, particularly in the high-risk population of HIV-infected patients presenting with focal neurologic signs.

Several questions are raised by these observations. The reasons for LP refusal are likely multifactorial, driven both by patient and practitioner factors. The consequences of not performing LPs in resource-limited settings are substantial. Without access to the diagnostic information provided by LPs, clinicians are unable to target therapeutics. Without the ability to use LPs as a therapeutic tool in conditions such as cryptococcal meningitis, many patients die due to refractory elevated ICP. A recent study has shown that in HIV-positive patients with cryptococcal meningitis, at least 1 therapeutic LP conducted in the first 10 days after diagnosis reduces mortality by 69%.¹⁰ In the COAT trial, a written script combined with a visual poster reduced LP refusal to <1% (figure). Further study regarding this important clinical issue of acceptance of LPs and whether such educational campaigns are replicable are needed. Qualitative studies of provider and patient perspectives on LPs and implementation of educational training programs are needed. Criteria for performance of LPs in regions worldwide where neuroimaging is not readily accessible are needed, as the potential diagnostic and therapeutic benefits of LPs may outweigh the small risk of herniation. Though these are basic measures, these tactics are likely to lead to improved targeted treatments and improved outcomes for many patients in sub-Saharan Africa and elsewhere.

Figure. Visual poster used in the Cryptococcal Optimal Antiretroviral Therapy Timing (COAT) trial for patient education regarding lumbar punctures.

Created by Kosuke Yasukawa. Reprinted with permission.

AUTHOR CONTRIBUTIONS

Kiran T. Thakur: design, writing, and editing of manuscript. Kondwelani Mateyo: editing of manuscript. Lottie Hachaambwa: editing of manuscript. Violet Kayamba: editing of manuscript. Macpherson Mallewa: editing of manuscript. Jane Mallewa: editing of manuscript. Ernest O. Nwazor: editing of manuscript. Tope Lawal: editing of manuscript. Chindo B. Mallum: editing of manuscript. Masharip Atadzhanov: editing of manuscript. David R. Boulware: design and editing of manuscript. Gretchen L. Birbeck: design and editing of manuscript. Omar K. Siddiqi: design, writing, and editing of manuscript.

STUDY FUNDING

No targeted funding reported.

DISCLOSURE

K. Thakur, K. Mateyo, L. Hachaambwa, V. Kayamba, M. Mallewa, E. Nwazor, T. Lawal, C. Mallum, M. Atadzhanov, and D. Boulware report no disclosures relevant to the manuscript. G. Birbeck reports NIH funding for research that includes data gained through LPs. O. Siddiqi reports no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.