*Department of Internal Medicine, Dodoma University, Dodoma, Tanzania;
†Department of Internal Medicine, Bugando Medical Centre, Mwanza, Tanzania;
‡Department of Internal Medicine, Weill Bugando School of Medicine, Mwanza, Tanzania;
§Department of Medicine, Center for Global Health, Weill Cornell Medical College, New York, NY;
‖Department of Internal Medicine, Muhimbili National Hospital, Dar-es-Salaam, Tanzania; and
Departments of ¶Microbiology;
#Biochemistry, Catholic University of Health and Allied Sciences—Bugando, Mwanza, Tanzania.
Correspondence to: Robert N. Peck, MD, Department of Medicine, Weill Bugando School of Medicine, PO Box 5034, Mwanza, Tanzania (e-mail: firstname.lastname@example.org).
Supported by a research grant from the University of Dodoma College of Health Sciences to J.M. and by grants from the National Institutes of Health Fogarty Foundation (TW000018) and the National Institute of Allergy and Infectious Diseases (K24 AI098627).
The authors have no conflicts of interest to disclose.
Received November 26, 2013
Accepted January 31, 2014
Cryptococcal meningitis (CM) is one of the most common severe opportunistic infections among adults with HIV worldwide and is responsible for more than 500,000 deaths per year in sub-Saharan Africa alone.1,2 In sub-Saharan Africa, CM accounts for 13% to 44% of all deaths of HIV-infected adults.3–5 Even after the rollout of antiretroviral therapy, the incidence of CM remains high at ∼3% per year among HIV-infected adults in sub-Saharan Africa.1,6
Although advances in the treatment of CM have decreased mortality in high-income countries, mortality as a result of CM in low-income countries remains high.1,7,8 The mortality rate from CM in sub-Saharan Africa has been estimated at 70% compared with 55% in other low- and middle-income countries and 10%–20% in high-income countries.1,9,10 This higher mortality is related to delayed diagnosis of both HIV and CM, inaccessibility of first-line treatment with combination amphotericin/flucytosine induction chemotherapy, and challenges in intensive intracranial pressure (ICP) management.7,11 In resource-poor settings, intensive ICP management is often not performed because of the unavailability of manometers and impracticality of ICP management protocols.
A previous study at our own hospital revealed that CM was the most common cause of death among HIV-infected adult inpatients and that the mortality rate of CM was ∼70%.12 In response to these findings, our hospital initiated an ongoing, continuous quality improvement project aimed to reduce mortality among adults admitted with CM. One aspect of this project was implementation of a protocol for intensive ICP management according to a strict schedule of serial lumbar punctures. The hospital decided to use widely available intravenous (IV) tubing sets for ICP measurement as has been previously recommended,13 but, to our knowledge, it was not rigorously implemented. Our hospital team predicted that, using this protocol, the mortality related to CM would be reduced by 30% compared with historical controls. This operational research study was conducted to describe the impact of this new intensive ICP management protocol and to compare ICP measurements using IV tubing sets with those with standard manual manometers.
This was an operational study in which we followed up outcomes of patients admitted with CM to the inpatient medical wards at Bugando Medical Centre (BMC) in Tanzania's Lake Victoria region between June 2011 and March 2012. BMC is a zonal hospital serving approximately 13 million people in northwestern Tanzania.
Hospital Meningitis Protocol and CM Registry
According to our hospital protocols, a diagnostic lumbar puncture is performed at the time of admission on all medical patients with ≥2/4 of the classic clinical features of meningitis including headache, fever, meningismus, and altered mental status (Glasgow coma scale <15). Patients with focal neurological deficits or unilateral papilledema do not undergo lumbar puncture because of concern for space-occupying lesions and the risk of herniation after lumbar puncture in our setting in which computed tomography scans are not consistently available. For all other patients, cerebrospinal fluid (CSF) is sent to the laboratory for standard investigations including Gram stain, India ink stain, bacterial culture, and cryptococcal antigen. Empiric treatment for bacterial meningitis (ceftriaxone 2 g IV daily for 2 weeks) is given to all patients unless Gram stains and bacterial cultures return negative. Other antibiotic therapy is added as indicated by investigations or clinician judgment.
For the past 3 years, all adults (age ≥18 years) admitted to the medical ward at BMC with the first episode of confirmed CM during the study period were considered for inclusion into a prospective patient registry. CM was confirmed with CSF cryptococcal antigen in all cases. After June 2011, the hospital decided to expand the data included in its registry of CM patients to incorporate clinical and laboratory parameters routinely collected for the management of CM patients including sociodemographic factors, HIV status, CD4+ T-cell count, antiretroviral therapy (ART) status, baseline symptoms and signs, CSF cryptococcal antigen titers, CSF India ink result, and baseline and subsequent CSF opening pressures by both manometer and IV giving set. In-hospital mortality was recorded at 2 weeks and at the time of discharge. Contact information was also obtained from the patient and at least 2 relatives of patients who survived to discharge so that 1-month mortality could be assessed. Beginning in June 2011, written informed consent was obtained before inclusion in the registry.
New Protocol for Intensive ICP Management
Because of poor persistently high CM mortality rates despite aggressive screening and treatment with high-dose IV fluconazole,12 in June 2011, our hospital decided to initiate a new protocol for intensive ICP management. Before June 2011, ICP management for patients with CM was left to the judgment of individual clinicians, but the new protocol (made available on all hospital wards) clearly indicated that (1) the ICP should be measured at the time of admission on all patients with suspected CM and (2) if the ICP was raised (>20 cm H2O), up to 30 mL of CSF were to be drained so as to achieve a closing pressure of ≤20 cm H2O or 50% of opening pressure according to expert recommendations.14,15 If the diagnosis of CM was confirmed, serial lumbar punctures to measure ICP were to be performed on hospital days 0, 3, 7, and 14 according to recent recommendations.14,16 Any patient with increased ICP at any time received daily therapeutic lumbar punctures (as above) until the ICP was ≤20 cm H2O and then resumed the normal schedule of serial lumbar punctures as above. No other changes were made to our hospital protocols for the diagnosis and management of meningitis during the study period or in the 6 months previously. Verbal informed consent was reconfirmed before all procedures.
To prove that ICP measurement with IV giving sets was reliable, our hospital decided that measurement of ICP should be done using both a manometer (Rocket spinal manometers; Rocket Medical, Washington, United Kingdom) and an IV giving set for the first year. Measurement of ICP by IV giving set was done by measuring the vertical height in centimeters reached by column of CSF rising in an IV giving set attached to the end of a spinal needle and held vertically. Measurement of ICP by IV giving set was always done after measurement with manometer, and the clinician who made this measurement was blinded to the manometer result according to the hospital protocol.
Clinical Evaluation and Follow-up
In hospitals such as BMC in which amphotericin B and flucytosine are not available, Tanzanian national guidelines recommend fluconazole for treatment of CM.17 In accordance with these guidelines, all patients with confirmed CM, including the historical controls, were treated with high-dose IV fluconazole (1200 mg/d) for 2 weeks during the induction phase and then 400 mg/d for 10 weeks. After completion of this 12-week course, patients remained on fluconazole 200 mg/d orally for secondary prophylaxis. At discharge, HIV-infected patients were referred for ongoing outpatient care at the BMC HIV clinic.
Data were entered using Microsoft Excel 2007 and analyzed using STATA IC/11.1 (StataCorp LP, College Station, TX). Continuous variables were summarized by median (interquartile ranges [IQRs]) or mean (SD) and categorical variables were summarized by proportions (percentages). The primary endpoint was 30-day mortality compared with historical controls. Required sample size before analysis was obtained by estimation of the sample size for a 2-sample comparison of proportions estimating a 30% reduction in mortality after the intervention. We used survival analysis Cox proportional hazards model and Kaplan–Meier survival curves to compare the in-hospital mortality rates between patients of this cohort who received intensive ICP management according to the new schedule and patients with confirmed CM who had been admitted to BMC in the previous 2 years before this new ICP management protocol had been implemented (historical controls). Bivariate logistic regression was used to determine the baseline characteristics associated with in-hospital mortality. We used Bland–Altman plots to determine the agreement between ICP measurement between manometer and IV giving set. In all analyses, significance level was set at 0.05.
Ethical approval was obtained from the joint Catholic University of Health and Allied Sciences (CUHAS) and BMC ethics review board as well as from the Institutional Review Board of Weill Cornell Medical College. All written and verbal informed consents were obtained from each patient or their next of kin (when the patient could not provide informed consent).
Between June 2011 and March 2012, 37 adults with CM were admitted to the inpatient medical wards at BMC. Of these, 2 patients were excluded because of recurrent or relapsed CM. Therefore, a total of 35 HIV-infected adults with a first episode of CM were enrolled and all of them agreed to both the initial and subsequent lumbar punctures. Sixty-four patients with CM who had been admitted to BMC in the 2 years before the initiation of the new ICP management protocol were included as historical controls.
Table 1 describes the baseline characteristics of study participants. The median (IQR) age of patients was 34 years (31–40). Slightly more than half of the patients (54%) were men, and 16 (46%) were farmers. Among these 35 HIV-infected adults with CM, 25 (71%) were aware of their HIV status on admission but only 12 (34%) were receiving ART. The median CD4 count was 46 cells per microliter (23–86) and 28 patients (80%) had CD4 counts less than 100 cells per microliter. Almost all patients (33/35, 94%) had headache and 22/35 (63%) had altered mental status on admission. Among these 35 adults with a positive CSF cryptococcal antigen, 29 (83%) had positive India ink tests and 32 (91%) had cryptococcal antigen titers of ≥1:64 in the CSF.
Progression of ICP and Symptoms
Figure 1 describes the trend in mean ICP and the proportion of patients with normal ICP (<20 cm H2O). Thirty-two patients (91%) had raised ICP with CSF opening pressure of ≥20 cm H2O at the time of admission. Baseline median CSF opening pressure (on admission) by manometer and IV giving set were 32 cm H2O (24–37) and 30 cm H2O (23–35), respectively. CSF opening pressures decreased significantly over time, from a median of 32 cm H2O (24–37) at baseline to 22 cm H2O (20.5–25) by day 14 (P for trend = 0.007). Of note, although the opening pressures did decline, even at day 14, the majority of patients (55%) still had abnormally elevated opening pressures. Despite this, the patients experienced significant improvement in symptoms over the first 2 weeks of hospitalization; by day 14, 18 (90%), 17 (85%), and 18 (90%) patients had no headache, fever, or altered mental status, respectively (P < 0.001 for each symptom).
Comparison of ICP Measurements by Manometer and IV Giving Set
The correlation between ICP measurement by manometer and IV giving set was excellent (r2 = 0.96). A Bland–Altman plot (Fig. 2) also revealed good agreement. These differences held constant over time and between patients.
Mortality Rate Compared With Historical Controls
Of the 35 adults with CM who were enrolled in our study, 15 of 35 patients (43%) died during their first admission. The remaining 20 of 35 patients (57%) were discharged home, after completing 2 weeks of high-dose IV fluconazole, with oral fluconazole 400 mg once daily and were followed until 30 days after the day of admission. Among these discharged patients, 1 of 20 (5%) developed severe headache and died at home.
In the 2 years before the implementation of this new ICP management protocol, 64 HIV-infected adult inpatients with CM were treated with the same regimen of IV fluconazole but with ICP management left to the personal judgment of the clinicians. Of those 64 patients, 48 of 64 patients (75%) died within 30 days of admission at a mortality rate of 75%. Thus, the overall 30-day mortality was significantly less in the patients who received intensive ICP management according to a strict protocol of serial lumbar punctures (16/35 vs. 48/64 historical controls; hazard ratio = 2.1; 95% CI: 1.1 to 3.8; P = 0.018) (Fig. 3).
Baseline Factors Associated With Mortality
When the ICP was analyzed in categories of 5 cm H2O (ie, <20, 21–25, 26–30, 31–35, 36–40), the odds ratio for mortality was 1.8 per 5 cm increase (95% CI: 1.0-3.2, P = 0.043). No other baseline characteristics were significantly associated with mortality including sex, age, occupation, baseline CSF parameters (India ink, cryptococcal antigen titers), CD4 count, or presence of headache, fever, photophobia, or altered mental status at baseline.
In this operational research study conducted in a Tanzanian hospital, intensive ICP management with serial lumbar punctures according to a strict protocol was associated with a significant reduction in 30-day mortality compared with historical controls (46% vs. 75%). Given that more than 500,000 patients die annually from CM in sub-Saharan Africa and that the majority of CM patients have elevated ICPs, routine implementation of this serial lumbar puncture protocol has the potential to save hundreds or even thousands of lives each year.
This study also provides quantitative verification of the accuracy and utility of IV giving sets for providing an accurate measurement of ICP. IV giving sets are widely available in hospitals in sub-Saharan Africa, where manometers are much more difficult to obtain. For resource-limited settings, the use of IV giving sets to measure ICP has previously been recommended,13 but to the best of our knowledge, it has not been systematically evaluated. In this study where >100 measurements of ICP were made independently by both manometer and IV giving set, the agreement between these 2 measurements was excellent. As a result of this study, our hospital has continued to use IV giving sets for ICP measurement when manometers are not available.
The majority of patients in this study (91%) had elevated ICP at admission. More than half had more elevated ICPs >30 cm H2O, and more than half had persistently-elevated ICPs at day 14 despite serial lumbar punctures. Our cohort's high burden of severe cryptococcal disease, marked not only by elevated ICP but also by high cryptococcal antigen titers, low CD4 counts, and frequent altered mental status, mirrors the severity of the disease reported in other studies from sub-Saharan Africa.14,18 Patients with these findings are at risk for increased mortality emphasizing even more the importance of our intervention in reducing mortality. Furthermore, our work confirms findings from other published studies that serial lumbar punctures contribute significantly to reduced mortality and improved short-term survival.10,14
Of note, 25% of our patients had been started on ART less than 3 months before their CM diagnosis, raising the possibility of unmasking immune reconstitution inflammatory syndrome. CM is thought to be the second most common cause of immune reconstitution inflammatory syndrome in sub-Saharan Africa after tuberculosis19,20 and is often associated with raised ICP and poor outcomes. Thus, our study highlights not only the urgency of early diagnosis to prevent advanced CM but, even better, of striving to prevent CM altogether by incorporating routine screening for cryptococcal antigenemia into outpatient HIV care.19,20
Despite the significant reduction in mortality associated with our hospital's new intensive ICP management protocol, the 45% mortality observed in the study group compared with mortality rates of ∼10%–20% achieved in developed countries indicates room for improvement.1,9,10 Importantly, both historical controls and patients in the current study received high-dose fluconazole monotherapy according to World Health Organization and Tanzanian national guidelines for CM treatment in settings in which amphotericin B and flucytosine are unavailable.21 Our work underscores recent calls to improve access to first-line treatment agents for CM in low- and middle-income countries in which the burden of disease is highest.22,23 Cost should not be an overwhelming obstacle to making these drugs available since the 2012 international wholesale price of 1200 mg of IV fluconazole ($19.08–25.92) is actually greater than the cost of 50 mg of IV amphotericin ($5.57–12.10).24
In our study, all 35 adult inpatients with CM (or their next of kin) consented to and accepted serial lumbar punctures for intensive ICP management. This difference may be partially explained by the fact that, in our cohort, more than half of patients had altered mental status at the time of admission and therefore consent was provided by their next of kin after the benefits of serial lumbar punctures were explained. Our work suggests that implementation of this life-saving protocol should not be hampered by concerns about acceptability of therapeutic lumbar punctures to patients and their families.
In conclusion, our operational research study conducted at a Tanzanian hospital suggests that mortality due to CM can be significantly reduced through implementation of serial lumbar punctures according to a strict protocol. In addition, utilization of low-cost, widely available IV giving sets provides reliable measurements of ICP and could allow for implementation of this protocol for intensive ICP management even in settings where manometers are not available. As part of our ongoing, continuous quality improvement project we are currently evaluating the benefit of short courses of amphotericin to further reduce the mortality rates in adults with CM admitted to our hospital.
The authors thank Mr. Zacharia Igembe for his assistance in the laboratory. The authors also thank the Director General of BMC and other members of the faculty of the BMC Department of Internal Medicine for their support.
1. Park BJ, Wannemuehler KA, Marston BJ, et al.. Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS. 2009;23:525–530.
2. Holmes CB, Losina E, Walensky RP, et al.. Review of human immunodeficiency virus type 1-related opportunistic infections in sub-Saharan Africa. Clin Infect Dis. 2003;36:652–662.
3. Mallory KF, Churchyard GJ, Kleinschmidt I, et al.. The impact of HIV infection on recurrence of tuberculosis in South African gold miners. Int J Tuberc Lung Dis. 2000;4:455–462.
4. Okongo M, Morgan D, Mayanja B, et al.. Causes of death in a rural, population-based human immunodeficiency virus type 1 (HIV-1) natural history cohort in Uganda. Int J Epidemiol. 1998;27:698–702.
5. French N, Gray K, Watera C, et al.. Cryptococcal infection in a cohort of HIV-1-infected Ugandan adults. AIDS. 2002;16:1031–1038.
6. Jarvis JN, Boulle A, Loyse A, et al.. High ongoing burden of cryptococcal disease in Africa despite antiretroviral roll out. AIDS. 2009;23:1182–1183.
7. Sloan DJ, Dedicoat MJ, Lalloo DG. Treatment of cryptococcal meningitis in resource limited settings. Curr Opin Infect Dis. 2009;22:455–463.
8. Leimann BCQ, Koifman RJ. Cryptococcal meningitis in Rio de Janeiro State, Brazil, 1994-2004. Cad Saude Publica. 2008;24:2582–2592.
9. Dromer F, Mathoulin-Pélissier S, Launay O, et al.. Determinants of disease presentation and outcome during cryptococcosis: the CryptoA/D study. PLoS Med. 2007;4:e21.
10. Van der Horst CM, Saag MS, Cloud GA, et al.. Treatment of cryptococcal meningitis associated with the acquired immunodeficiency syndrome. National Institute of Allergy and Infectious Diseases Mycoses Study Group and AIDS Clinical Trials Group. N Engl J Med. 1997;337:15–21.
11. Sloan D, Dlamini S, Paul N, et al.. Treatment of acute cryptococcal meningitis in HIV infected adults, with an emphasis on resource-limited settings. Cochrane Database Syst Rev. 2008;4:CD005647.
12. Wajanga BM, Kalluvya S, Downs JA, et al.. Universal screening of Tanzanian HIV-infected adult inpatients with the serum cryptococcal antigen to improve diagnosis and reduce mortality: an operational study. J Int AIDS Soc. 2011;14:48.
13. Jackson A, Hosseinipour MC. Management of cryptococcal meningitis in sub-saharan Africa. Curr HIV/AIDS Rep. 2010;7:134–142.
14. Bicanic T, Brouwer AE, Meintjes G, et al.. Relationship of cerebrospinal fluid pressure, fungal burden and outcome in patients with cryptococcal meningitis undergoing serial lumbar punctures. AIDS. 2009;23:701–706.
15. Graybill JR, Sobel J, Saag M, et al.. Diagnosis and management of increased intracranial pressure in patients with AIDS and cryptococcal meningitis. The NIAID Mycoses Study Group and AIDS Cooperative Treatment Groups. Clin Infect Dis. 2000;30:47–54.
16. Fessler RD, Sobel J, Guyot L, et al.. Management of elevated intracranial pressure in patients with Cryptococcal meningitis. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;17:137–142.
17. The United Republic of Tanzania Ministry of Health and Social Welfare. National Guideline for the Management of HIV and AIDS. 4th ed. Dar es Salaam, Tanzania: 2012.
18. Nussbaum JC, Jackson A, Namarika D, et al.. Combination flucytosine and high-dose fluconazole compared with fluconazole monotherapy for the treatment of cryptococcal meningitis: a randomized trial in Malawi. Clin Infect Dis. 2010;50:338–344.
19. Jarvis JN, Govender N, Chiller T, et al.. Cryptococcal antigen screening and preemptive therapy in patients initiating antiretroviral therapy in resource-limited settings: a proposed algorithm for clinical implementation. J Int Assoc Physicians AIDS Care (Chic). 2012;11:374–379.
20. Rajasingham R, Meya DB, Boulware DR. Integrating cryptococcal antigen screening and pre-emptive treatment into routine HIV care. J Acquir Immune Defic Syndr. 2012;59:e85–e91.
21. World Health Organization. Rapid Advice Diagnosis, Prevention and Management of Cryptococcal Disease in HIV-infected Adults, Adolescents and Children. Geneva, Switzerland: 2011.
22. Loyse A, Dromer F, Day J, et al.. Flucytosine and cryptococcosis: time to urgently address the worldwide accessibility of a 50-year-old antifungal. J Antimicrob Chemother. 2013;68:2435–2444.
23. Loyse A, Thangaraj H, Easterbrook P, et al.. Cryptococcal meningitis: improving access to essential antifungal medicines in resource-poor countries. Lancet Infect Dis. 2013;13:629–637.
24. Management Sciences for Health. International Drug Price Indicator Guide (2012). Available at: http://erc.msh.org/dmpguide
. Accessed December 16, 2013.