 |
|
|
| |
Bacterial Pneumonia Incidence
Drops Sharply in Swiss HIV Cohort Study
|
| |
| |
18th European AIDS Conference, EACS 2021, October 27-30, 2021, London
Mark Mascolini
New diagnoses of bacterial pneumonia fell from 12.8 cases per 1000 person-years in 2008-2009 to 8.1 cases per 1000 person-years in 2017-2018 in a 9213-person analysis of the Swiss HIV Cohort Study (SHCS) [1]. Compared with a CD4 count above 500, even a count of 350 to 499 boosted risk of bacterial pneumonia. Other factors linked to a higher bacterial pneumonia risk were viral load above 200 copies, obstructive airway disease, smoking, and use of proton pump inhibitors.
Bacterial pneumonia continues to send many people with HIV to the hospital, but factors explaining this high pneumonia rate are poorly defined in the current antiretroviral era. To address this knowledge gap, SHCS investigators set out to estimate bacterial pneumonia incidence in people with HIV using a time-to-survival analysis, to identify pneumonia risk factors using marginal models, and to assess the relationship between risk factors and changes in pneumonia incidence using multiple mediation analysis.
The study involved HIV-positive people enrolled in the SHCS between August 1, 2008 and December 31, 2018. Researchers assessed four risk factors at cohort enrollment: ethnicity, education level, likely source of HIV infection, and pack-years smoking. Time-updated variables were age, smoking status, CD4 count, viral load, prior pneumonia, proton pump inhibitors, angiotensin-converting-enzyme (ACE) inhibitors, obstructive airway disorders, kidney disease, and pneumococcal vaccination.
The analysis focused on 9213 SHCS members with no missing values needed for the analysis. Of those 9213 people, 233 (2.5%) had 272 cases of pneumonia during 28,960 person-years of follow-up. People with a bacterial pneumonia diagnosis differed from those without pneumonia in gender (68.2% vs 72.8% male), white ethnicity (87.1% vs 76.6%), bachelor's degree or higher education (18.4% vs 35.5%), nadir CD4 count (392.5 vs 479), smoking one or more cigarettes daily (68.2% vs 44.5%), proton pump inhibitor use (35.6% vs 16.7%), ACE inhibitor use (24.9% vs 12.2%), kidney disease (23.2% vs 17.0%), obstructive airway disease (20.6% vs 5.9%), ever had nonconjugated pneumococcal vaccine (54.9% vs 42.2%), ever had conjugated pneumococcal vaccine (51.1% vs 45.2%).
From 2008-2009 to 2017-2018, incident (newly diagnosed) bacterial pneumonia fell by one third in this HIV group from 12.8 to 8.1 cases per 1000 person-years, a highly significant drop (P = 0.001). One factor independently lowered the risk of bacterial pneumonia: education to at least a bachelor's degree (hazard ratio [HR] 0.63).
Ten factors independently predicted a greater risk of bacterial pneumonia: every additional 10 years of age (HR 1.39), injection drug use as the HIV risk factor in women (HR 2.28), viral load above 200 copies (vs below) (HR 1.92), CD4 count 350 to 499 (vs 500 or more) (HR 1.39), CD4 count 200 to 249 (HR 1.5), CD4 count 50 to 199 (HR 2.46), CD4 count below 50 (HR 7.68), smoking 1 or more cigarettes daily (HR 1.46), proton pump inhibitor use (HR 1.66), obstructive airway disease (HR 2.24), and any prior pneumonia event (HR 4.82). All these variables were time-updated except injection drug use as an HIV risk factor in women.
The SHCS team performed a multiple mediation analysis to figure whether changing risk factors over time explained the decreasing incidence of bacterial pneumonia during the study period. This analysis determined that declining bacterial pneumonia diagnoses over the last decade can be explained by the falling proportion of cohort members with a CD4 below 500, a viral load above 200 copies, and smoking one or more cigarettes daily.
The researchers proposed that the improving HIV care cascade led to higher CD4 counts and lower viral loads, and those changes-along with reduced smoking by people with HIV-"mediated a substantial decrease in bacterial pneumonia incidence."
A decade ago in the United States, bacterial pneumonia incidence was about 35-fold higher in people with than without HIV [2]. In a review published at that time, US researchers estimated that incidence in HIV-positive people had already plunged from about 5 to 20 cases per 100 person-years to about 1 to 5 cases per 100 person-years [2]. An incidence of 1 case per 100 person years is similar to the 8 cases per 1000 person-years found in 2017-2018 in Switzerland.
References
1. Balakrishna S, Wolfensberger A, Kachalov V, et al. Decreasing incidence and determinants of bacterial pneumonia in people living with HIV: the Swiss HIV cohort study. 18th European AIDS Conference, EACS 2021, October 27-30, 2021, London. Abstract OS2/1.
2. Segal LN, Methé BA, Nolan A, et al. HIV-1 and bacterial pneumonia in the era of antiretroviral therapy. Proc Am Thorac Soc. 2011;8:282-287. doi: 10.1513/pats.201006-044WR.
|
| |
|
|
|
|
|
