Delays in lumbar puncture are independently associated with mortality in cryptococcal meningitis: a nationwide study

Abstract

Background

Cryptococcal meningitis (CM) is the most serious presentation of invasive cryptococcosis. Seen in patients with and without HIV infection, CM is associated with significant morbidity and mortality. Early lumbar puncture is a cornerstone of treatment in cryptococcal meningitis. We present findings from a nationwide analysis of patients admitted with CM in the United States between 2007 and 2016, with the aim of determining the impact of delays in lumbar puncture on inpatient outcomes.

Methods

The national inpatient sample (NIS) database was queried for all inpatient visits for cryptococcal meningitis between January 2007 and December 2016. Logistic regression models were used to determine risk factors for inpatient mortality, prolonged admissions, and delays in obtaining an initial lumbar puncture.

Results

The annual number of admissions for CM decreased during the study interval, from 3590 in 2007 to 2830 in 2016. Mortality did not change over this period (9.9%); however, length of stay and inpatient cost significantly increased. The proportion of patients with HIV declined from 70.7% to 54.0%. Delay in lumbar puncture beyond the first 24 h was independently associated with mortality (OR = 1.55, CI = 1.31–1.82, p-value <.001). Patients admitted on a weekend, those of African–American ethnicity, and those without a known history of HIV were more likely to have delays in obtaining an early LP. HIV patients had a lower risk of mortality (OR = 0.77, CI = 0.68–0.86, p-value <.001).

Conclusion

We found an independent association of delay in early lumbar puncture with worsened patient outcomes. Inpatient mortality for patients with CM continues to remain high, with an increasing proportion of patients without underlying HIV infection. There were significant deviations in management of CM from Infectious Diseases Society of America (IDSA) guidelines.

Keywords:

• Cryptococcal meningitis
• HIV
• large database analysis
• lumber puncture
• inpatient outcomes

Disclosure statement

No potential conflict of interest was reported by the author(s).