Diagnostic test accuracy of jolt accentuation for headache in acute meningitis in the emergency setting

Abstract

Background Meningitis is inflammation of the meninges, the layers that protect the brain and spinal cord. Acute meningitis is an emergent disease that develops over the course of hours to several days. Delay in treatment can lead to serious outcomes. Inflammation of the meninges is assessed by analysing cerebrospinal fluid. Identifying the pathogen in cerebrospinal fluid is another way to diagnose meningitis. Cerebrospinal fluid is collected by doing a lumbar puncture, which is an invasive test, and can be avoided if a physical examination excludes the diagnosis of meningitis. However, most physical examinations, such as nuchal rigidity, Kernig's test, and Brudzinski's test, are not sufficiently sensitive to exclude meningitis completely. Jolt accentuation of headache is a new and less well-recognised physical examination, which assesses meningeal irritation. It is judged as positive if the headache is exacerbated by rotating the head horizontally two or three times per second. A 1991 observational study initially reported high sensitivity of this examination to predict pleocytosis. Pleocytosis, an abnormally high cerebrospinal fluid sample white cell count, is an accepted indicator of nervous system infection or inflammation. Jolt accentuation of headache may therefore accurately rule out meningitis without the use of lumbar puncture. However, more recent cross-sectional studies have reported variable diagnostic accuracy. Objectives To estimate the diagnostic accuracy of jolt accentuation of headache for detecting acute meningitis in emergency settings. Secondary objectives: to investigate the sources of heterogeneity, including study population, patient condition, and types of meningitis. Search methods We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (Ovid), and Embase (Elsevier) to 27 April 2020. We searched ClinicalTrials.gov, the World Health Organization International Clinical Trials Registry Platform, and Ichushi-Web Version 5.0 to 28 April 2020. Selection criteria We included cross-sectional studies that assessed the diagnostic accuracy of jolt accentuation of headache for people with suspected meningitis in emergency settings. We included participants of any age and any severity of illness. Meningitis should be diagnosed with any reference standard, such as cerebrospinal fluid pleocytosis, proof of causative agents, or autopsy. Data collection and analysis Two review authors independently collated study data. We assessed methodological quality of studies using QUADAS-2 criteria. We used a bivariate random-effects model to determine summary estimates of sensitivity and specificity where meta-analysis was possible. We performed sensitivity analyses to validate the robustness of outcomes. We assessed the certainty of the evidence using the GRADE approach. Main results We included nine studies (1161 participants). Five studies included only adults. Four studies included both adults and children; however, the proportion was not reported in three of these studies. The youngest child reported in the studies was aged 13 years. There was no study including only children. The reference standard was pleocytosis in eight studies, and the combination of pleocytosis and increased protein in the cerebrospinal fluid in one study. Two studies also used smear or positive culture of cerebrospinal fluid. Risk of bias and concern about applicability was high in the participant selection domain for all included studies and the consciousness subgroup. Overall, pooled sensitivity was 65.3% (95% confidence interval (CI) 37.3 to 85.6), and pooled specificity was 70.4% (95% CI 47.7 to 86.1) (very low-certainty evidence). We established the possibility of heterogeneity from visual inspection of forest plots. However, we were unable to conduct further analysis for study population, types of meningitis, and participants' condition, other than disturbance of consciousness (a secondary outcome). Amongst participants whose consciousness was undisturbed (8 studies, 921 participants), pooled sensitivity and specificity were 75.2% (95% CI 54.3 to 88.6) and 60.8% (95% CI 43.4 to 75.9), respectively (very low-certainty evidence). Authors' conclusions Jolt accentuation for headache may exclude diagnoses of meningitis in emergency settings, but high-quality evidence to support use of this test is lacking. Even where jolt accentuation of headache is negative, there is still the possibility of acute meningitis. This review identified the possibility of heterogeneity. However, factors that contribute to heterogeneity are incompletely understood, and should be considered in future research.