Abstract
Background
Pneumocystis pneumonia (PCP) is a life-threatening but treatable and preventable fungal infection in immunocompromised persons. Previous studies suggest that persons without HIV who develop PCP (PCPHIV-) experience more acute, severe illness than persons with HIV who develop PCP (PCPHIV+). We analyzed health insurance claims data to compare demographics, underlying conditions, symptoms, and prescriptions for PCPHIV+ and PCPHIV-.
Methods
We used the IBM MarketScan Research Databases to identify patients diagnosed with PCP during 2011–2015. We analyzed claims 1 year before to 3 months after diagnosis to compare PCPHIV+ and PCPHIV-.
Results
Among 3938 patients, 70.4% were PCPHIV-. Compared with PCPHIV+, PCPHIV- were more likely to be older (median, 60 vs 45 years; P < .0001), female (51.5% vs 20.2%; P < .0001), hypoxemic (13.5% vs 7.1%; P < .0001), and to die within 90 days (6.6% vs 4.2%; P < .0001). The most common underlying conditions among PCPHIV- included chronic pulmonary diseases (54.6%), solid tumors (35.1%), hematologic malignancies (20.1%), and rheumatologic conditions (14.0%). The median time between the first visit for PCP-related symptoms and PCP diagnosis was longer for PCPHIV- than PCPHIV+ (25 vs 16 days; P < .0001). In the 3 months before PCP diagnosis, PCPHIV- were less likely to have an outpatient prescription for PCP prophylaxis than PCPHIV+ (6.9% vs 10.6%; P = .0001).
Conclusions
PCPHIV- may experience a prolonged illness course and diagnostic delays compared with PCPHIV+. Clinicians should maintain a high index of suspicion for PCP in immunocompromised patients with compatible symptoms, regardless of HIV status.
Keywords: diagnosis, hospitalization, immunocompromise, Pneumocystis pneumonia, prophylaxis
Pneumocystis pneumonia (PCP) is a life-threatening opportunistic infection caused by the fungus Pneumocystis jirovecii (formerly P. carinii) [1, 2]. PCP was rare before the HIV/AIDS epidemic began in the 1980s [3] but became well known as a common AIDS-defining illness [4]. The routine use of PCP prophylaxis with trimethoprim-sulfamethoxazole (TMP-SMX) and antiretroviral therapy in persons with HIV (PWH) led to a dramatic decrease in PCP incidence in the United States [5–8]. Currently, most persons in the United States who develop PCP are not immunocompromised by HIV, but rather from immunosuppressive medications and medical conditions like malignancies, solid organ transplants, rheumatological conditions, and primary immunodeficiency syndromes [9].
PCP is classically thought to have different clinical presentations in PWH and HIV-uninfected patients. PWH who develop PCP (PCPHIV+) present subacutely with progressive worsening of chills, cough, dyspnea, and chest pain over the course of several weeks [10, 11]. In contrast, previous research suggests that HIV-uninfected persons who develop PCP (PCPHIV-) present acutely over several days and have a higher likelihood of respiratory failure and death [11, 12]. Prophylaxis against PCP with TMP-SMX is effective both for PWH and HIV-uninfected persons [13, 14]. Although the use of PCP prophylaxis for PWH has been a well-established practice for decades [15], consensus is lacking about which HIV-uninfected patients should receive PCP prophylaxis [14].
Despite being treatable and preventable, PCP remains a serious and costly public health issue [9, 16–18]. PCP is increasingly described in non-HIV-immunocompromised patients, yet most studies of PCP have focused on PWH, and epidemiological data about PCP in HIV-uninfected patients are sparse. To provide an update on the epidemiology of PCP in the United States with the goal of informing clinical practice, we analyzed administrative data to compare demographics, underlying conditions, clinical presentations, and outpatient drug prescriptions in PCPHIV+ and PCPHIV-.
METHODS
We used data from the 2010–2016 IBM MarketScan Research Databases. The MarketScan Commercial Database and the Medicare Supplemental Database contain individual-level insurance claims data representing a range of health services, including hospitalizations, outpatient visits, and outpatient prescriptions for >245 million employees, dependents, and retirees throughout the United States. Because these databases are fully de-identified, our analysis did not require review by the Centers for Disease Control and Prevention institutional review board.
We identified patients with PCP using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), diagnosis code 136.3 and Tenth Revision (ICD-10-CM) diagnosis code B59 during 2011–2015. The index date was the date a PCP diagnosis code was first used in the study period. Of >89 million enrollees, 9727 patients had a PCP code during 2011–2015.
We required continuous enrollment during the year before to 3 months after the index date (n = 4509 patients) or continuous enrollment during the year before the index date and in-hospital death (n = 378). To attempt to identify incident cases of PCP, we subsequently excluded patients with a PCP code in the 3 months before the index date (n = 18). We also excluded patients with insurance plans that did not contribute prescription drug data to MarketScan (n = 886), for a final cohort of 3983 patients.
We used ICD codes to identify underlying conditions diagnosed on or within 12 months before the index date and to identify symptoms and clinical findings suggestive of PCP (fever, cough, dyspnea, and chest pain) in the 3 months before, on, or after the index date (Supplementary Table 1). We also examined non-Pneumocystis pneumonia codes before the index date to identify possible misdiagnosis. We defined chronic steroid use as an outpatient prescription for a ≥4-week supply of prednisone at a dose ≥20 milligrams daily, or the bioequivalent dose of another systemic corticosteroid drug, during the 3 months before the index date. Because treatment for PCP generally lasts ≤3 weeks, we defined PCP prophylaxis as having an outpatient prescription for a >21-day supply of a PCP prophylactic drug (TMP-SMX, atovaquone, dapsone, pentamidine, or clindamycin with primaquine) during the 3 months before the index date.
We performed descriptive analyses and examined differences between PCPHIV+ and PCPHIV- in terms of demographic features, diagnoses, hospitalization status, 90-day in-hospital mortality, and outpatient drug prescriptions using χ 2 or Fisher exact tests for categorical variables and Wilcoxon rank-sum tests for continuous variables.
RESULTS
In the final cohort of 3938 patients, 70.4% were PCPHIV- and 29.6% were PCPHIV+. PCPHIV- were older than PCPHIV+ (median, 60 vs 45 years; P < .0001), but 3.0% of PCPHIV- were <18 years old. Compared with PCPHIV+, PCPHIV- were more likely to be female (51.5% vs 20.2%; P < .0001). The most common US Census division of residence was South Atlantic for both PCPHIV- (19.2%) and PCPHIV+ (31.1%) (Table 1).
Table 1.
Demographic Features of Patients Diagnosed With Pneumocystis Pneumonia in a Commercially Insured US Population, 2011–2015
| All Patients (n = 3938) | HIV-Uninfected (n = 2773) | Persons With HIV (n = 1165) | |||||
|---|---|---|---|---|---|---|---|
| Characteristic | No. | (%) | No. | (%) | No. | (%) | P Value |
| Age group, y | <.0001 | ||||||
| 0–17 | 118 | (3.0) | 117 | (4.2) | a | a | |
| 18–34 | 410 | (10.4) | 180 | (6.5) | 230 | (19.7) | |
| 35–44 | 532 | (13.5) | 196 | (7.1) | 336 | (28.8) | |
| 45–54 | 868 | (22.0) | 461 | (16.6) | 407 | (34.9) | |
| 55–64 | 964 | (24.5) | 800 | (28.8) | 164 | (14.1) | |
| ≥65 | 1046 | (26.6) | 1019 | (36.7) | 27 | (2.3) | |
| Sex | <.0001 | ||||||
| Male | 2275 | (57.8) | 1345 | (48.5) | 930 | (79.8) | |
| Female | 1663 | (42.2) | 1428 | (51.5) | 235 | (20.2) | |
| US Census division of primary beneficiary’s residence | <.0001 | ||||||
| New England | 154 | (3.9) | 129 | (4.7) | 25 | (2.1) | |
| Mid-Atlantic | 627 | (15.9) | 470 | (16.9) | 157 | (13.5) | |
| East North Central | 650 | (16.5) | 518 | (18.7) | 132 | (11.3) | |
| West North Central | 131 | (3.3) | 99 | (3.6) | 32 | (2.7) | |
| South Atlantic | 894 | (22.7) | 532 | (19.2) | 362 | (31.1) | |
| East South Central | 197 | (5.0) | 134 | (4.8) | 63 | (5.4) | |
| West South Central | 511 | (13.0) | 301 | (10.9) | 210 | (18.0) | |
| Mountain | 237 | (6.0) | 190 | (6.9) | 47 | (4.0) | |
| Pacific | 483 | (12.3) | 369 | (13.3) | 114 | (9.8) | |
| Puerto Rico or US Virgin Islands | 54 | (1.4) | 31 | (1.1) | 23 | (2.0) | |
aData are not presented due to small sample size.
The most common underlying conditions among PCPHIV- were chronic pulmonary diseases (54.6%), solid tumors (35.1%), hematologic malignancies (20.1%), rheumatologic conditions (14.0%), and solid organ transplantation (6.6%) (Table 2). Less frequent conditions included primary immunodeficiency syndromes (2.7%), malnutrition (2.5%), inflammatory bowel disease (2.4%), and multiple sclerosis (0.6%). Over 99% of patients with malnutrition had an additional underlying condition identified. Among PCPHIV-, 16.8% had no underlying condition identified that might predispose them to PCP. Compared with PCPHIV+, PCPHIV- were less likely to be hospitalized for PCP (67.3% vs 81.4%; P < .0001) but more likely to die within 90 days while in the hospital (6.6% vs 4.2%; P = .0046). PCPHIV- with no identified underlying conditions were less likely to be hospitalized (44.8% vs 67.3%; P < .0001) or to die within 90 days (2.1% vs 6.6%; P < .0001) compared with other PCPHIV-. Among PCPHIV-, 90-day mortality was highest in patients with malnutrition (24.6%), primary immunodeficiency disorders (13.5%), hematologic malignancies (11.5%), and solid tumors (11.4%). Ninety-day in-hospital mortality was 7.2% for patients with rheumatologic conditions and 14.3% for patients with granulomatosis with polyangiitis. Patients with solid organ transplant had a similar 90-day mortality compared with PCPHIV+ (4.9% vs 4.2%; P = .7321).
Table 2.
Hospitalizations and 90-Day In-Hospital Mortality for US Pneumocystis Pneumonia Patients by Underlying Condition, 2011–2015a
| Hospitalized | Died | |||||||
|---|---|---|---|---|---|---|---|---|
| Conditionb | No. | (%) | No. | (%) | P Value | No. | (%) | P Value |
| All patients | 3938 | 2813 | (71.4) | 231 | (5.9) | |||
| Persons with HIV | 1165 | (29.6) | 948 | (81.4) | 49 | (4.2) | ||
| HIV-uninfected patients | 2773 | (70.4) | 1865 | (67.3) | <.0001 | 182 | (6.6) | .0046 |
| Chronic pulmonary disease | 1514 | (54.6) | 1011 | (66.8) | <.0001 | 91 | (6.0) | .1104 |
| Chronic obstructive pulmonary disease | 792 | (28.6) | 588 | (74.2) | 50 | (6.3) | ||
| Asthma | 476 | (17.2) | 316 | (66.4) | 29 | (6.1) | ||
| Bronchiectasis | 169 | (6.1) | 130 | (76.9) | 11 | (6.5) | ||
| Cystic fibrosis | 54 | (1.9) | 37 | (68.5) | c | c | ||
| Cryptogenic organizing pneumonia | 16 | (0.6) | 15 | (93.8) | c | c | ||
| Sarcoidosis | 64 | (2.3) | 41 | (64.1) | c | c | ||
| Pneumoconiosis | 86 | (3.1) | 71 | (82.6) | 7 | (8.1) | ||
| Idiopathic pulmonary fibrosis | 34 | (1.2) | 29 | (85.3) | c | c | ||
| Pulmonary fibrosis (not otherwise specified) | 803 | (29.0) | 536 | (66.7) | 40 | (5.0) | ||
| Solid tumor | 973 | (35.1) | 767 | (78.8) | .1929 | 111 | (11.4) | <.0001 |
| Hematologic malignancy | 557 | (20.1) | 484 | (86.9) | .0011 | 64 | (11.5) | <.0001 |
| Leukemia | 276 | (10.0) | 243 | (88.0) | 29 | (10.5) | ||
| Hodgkin’s lymphoma | 64 | (2.3) | 58 | (90.6) | 6 | (9.4) | ||
| Non-Hodgkin’s lymphoma | 322 | (11.6) | 279 | (86.6) | 43 | (13.4) | ||
| Multiple myeloma | 101 | (3.6) | 95 | (94.1) | 10 | (9.9) | ||
| Stem cell or bone marrow transplant | 169 | (6.1) | 153 | (90.5) | 17 | (10.1) | ||
| Rheumatologic conditions | 387 | (14.0) | 273 | (70.5) | <.0001 | 28 | (7.2) | .0481 |
| Rheumatoid arthritis | 231 | (8.3) | 164 | (71.0) | 15 | (6.5) | ||
| Systemic lupus erythematosus | 81 | (2.9) | 63 | (77.8) | 6 | (7.4) | ||
| Granulomatosis with polyangiitis/polyarteritis nodosa | 35 | (1.3) | 27 | (77.1) | 5 | (14.3) | ||
| Microscopic polyangiitis | 12 | (0.4) | 12 | (100.0) | c | c | ||
| Dermatomyositis | 42 | (1.5) | 29 | (69.0) | c | c | ||
| Polymyositis | 33 | (1.2) | 21 | (63.6) | c | c | ||
| Scleroderma | 35 | (1.3) | 19 | (54.3) | c | c | ||
| Polymyalgia rheumatica/giant cell arteritis | 27 | (1.0) | 18 | (66.7) | c | c | ||
| Solid organ transplant | 183 | (6.6) | 158 | (86.3) | .1863 | 9 | (4.9) | .7321 |
| Heart transplant | 16 | (0.6) | 15 | (93.8) | c | c | ||
| Lung transplant | 28 | (1.0) | 18 | (64.3) | c | c | ||
| Kidney transplant | 111 | (4.0) | 102 | (91.9) | 6 | (5.4) | ||
| Liver transplant | 37 | (1.3) | 35 | (94.6) | c | c | ||
| Primary immunodeficiency disorders | 74 | (2.7) | 68 | (91.9) | .0760 | 10 | (13.5) | .0074 |
| X-linked lymphoproliferative disease | 58 | (2.1) | 55 | (94.8) | 8 | (13.8) | ||
| Severe combined immunodeficiency | 9 | (0.3) | 8 | (88.9) | c | c | ||
| Other primary immunodeficiency disorders | 10 | (0.4) | 8 | (80.0) | c | c | ||
| Severe malnutrition | 69 | (2.5) | 67 | (97.1) | <.0001 | 17 | (24.6) | <.0001 |
| Inflammatory bowel disease | 66 | (2.4) | 57 | (86.4) | .1768 | 7 | (10.6) | .0442 |
| Crohn’s disease | 37 | (1.3) | 33 | (89.2) | c | c | ||
| Ulcerative colitis | 38 | (1.4) | 32 | (84.2) | 5 | (13.2) | ||
| Multiple sclerosis | 18 | (0.6) | 15 | (83.3) | >.999 | c | c | |
| No underlying condition identified | 466 | (16.8) | 209 | (44.8) | <.0001 | 10 | (2.1) | <.0001 |
a P values are based on comparison of percent hospitalized or 90-day in-hospital mortality with the HIV-infected group.
bPatients could have received diagnosis codes for multiple conditions.
cData are not presented due to small sample size.
In the 3 months before the index date, nearly 30% of all patients with PCP had 2 or more visits with PCP symptoms. PCPHIV- had more provider visits than PCPHIV+ for symptoms consistent with PCP (59.6% vs 53.5%; P = .0004) and were more likely to have ≥3 visits for PCP symptoms before the index date (18.4% vs 10.8%) (Table 3). The median time between the first visit for PCP symptoms and the index date was 25 days for PCPHIV- (mean [range], 31.2 [1–89]) and 16 days for PCPHIV+ (mean [range], 26.4 [1–89]; P < .0001) (Figure 1). Dyspnea was more common in PCPHIV- than in PCPHIV+ but was the most common symptom for both patient groups (38.2% vs 30.3%; P < .0001). Hypoxemia was nearly twice as common in PCPHIV- compared with PCPHIV+ (13.5% vs 7.1%; P < .0001). A similar proportion of PCPHIV- and PCPHIV+ were diagnosed with non-Pneumocystis pneumonia before the index date (26.7% vs 28.2%; P = .3462).
Table 3.
Symptoms, Clinical Findings, and Prescriptions in the 3 Months Before Pneumocystis Pneumonia Diagnosis, 2011–2015
| All Patients (n = 3938) | HIV-Uninfected (n = 2773) | Persons With HIV (n = 1165) | |||||
|---|---|---|---|---|---|---|---|
| Characteristic | No. | (%) | No. | (%) | No. | (%) | P Value |
| Symptoms suggestive of PCP | 2276 | (57.8) | 1653 | (59.6) | 623 | (53.5) | .0004 |
| Fever/chills | 635 | (16.1) | 439 | (15.8) | 196 | (16.8) | .4477 |
| Cough | 1083 | (27.5) | 725 | (26.1) | 358 | (30.7) | .0034 |
| Dyspnea | 1413 | (35.9) | 1060 | (38.2) | 353 | (30.3) | <.0001 |
| Chest pain | 608 | (15.4) | 454 | (16.4) | 154 | (13.2) | .0137 |
| Hypoxemia | 457 | (11.6) | 374 | (13.5) | 83 | (7.1) | <.0001 |
| No. of visits for symptoms consistent with PCP | |||||||
| 0 | 1822 | (46.3) | 1241 | (44.8) | 572 | (49.1) | <.0001 |
| 1 | 971 | (24.7) | 651 | (23.5) | 320 | (27.5) | |
| 2 | 518 | (13.2) | 371 | (13.4) | 147 | (12.6) | |
| ≥3 | 636 | (16.2) | 510 | (18.4) | 126 | (10.8) | |
| Mean, median duration between first symptom or clinical finding and index date (range), d | 31.1, 24.0 | (1–89) | 31.2, 25.0 | (1–89) | 26.4, 16.0 | (1–89) | <.0001 |
| Diagnosed with non-Pneumocystis pneumonia | 1068 | (27.1) | 740 | (26.7) | 328 | (28.2) | .3462 |
| Outpatient prescriptions | |||||||
| Chronic steroid usea | 415 | (10.5) | 402 | (14.5) | 13 | (1.1) | <.0001 |
| PCP prophylaxis | 314 | (8.0) | 190 | (6.9) | 124 | (10.6) | .0001 |
| Trimethoprim-sulfamethoxazole | 277 | (7.0) | 168 | (6.1) | 109 | (9.4) | .0002 |
| Other drugsb | 37 | (1.2) | 22 | (0.8) | 15 | (1.3) | .1423 |
| Systemic outpatient antibacterial medication, excluding drugs effective against PCP | 1625 | (41.3) | 1099 | (39.6) | 526 | (45.2) | .0014 |
Abbreviation: PCP, Pneumocystis pneumonia.
aChronic steroid use was defined as having an outpatient prescription for a ≥4-week supply of prednisone at a dose of ≥20 milligrams daily or the bioequivalent dose of another systemic corticosteroid drug during the 3 months before the index date.
bOther PCP prophylaxis drugs included atovaquone, pentamidine, dapsone, and primaquine with clindamycin.
Figure 1.
Scaled density plots depicting symptoms/diagnoses (A) and prescription medications (B) within 3 months before and after Pneumocystis pneumonia (PCP) diagnosis. In HIV-uninfected persons, codes for symptoms consistent with PCP (fever, cough, dyspnea, and/or chest pain), non-Pneumocystis pneumonia diagnoses, hypoxemia, and prescriptions for non-Pneumocystis-treating antibiotics increased in the days leading up to the index date and declined thereafter. Before the index date, more HIV-uninfected persons had outpatient prescriptions for corticosteroid drugs than persons with HIV (PWH), but prescriptions for corticosteroids rose sharply in PWH after the index date.
In the 3 months before the index date, PCPHIV- were more likely to be on chronic steroids compared with PCPHIV+ (14.5% vs 1.1%; P < .0001) but less likely to have a prescription for PCP prophylaxis (6.9% vs 10.6%; P = .0001). Forty-one percent of all PCP patients received an outpatient prescription for systemic antibacterial therapy other than PCP-treating drugs during the 3 months before the index date.
DISCUSSION
Our analysis of PCP patients in a large commercial health insurance claims database found that most patients were HIV-uninfected and PCPHIV- had greater diagnostic delays and mortality compared with PCPHIV+. More than half of PCPHIV- had provider visits with diagnostic codes for PCP symptoms in the 3 months before diagnosis, with PCP symptoms increasing in the days leading up to their index dates. Diagnoses for non-Pneumocystis pneumonia and prescriptions for non-PCP-treating antibiotics also increased in the days leading up to the index date, suggesting the possibility of misdiagnosis and inappropriate treatments. The prolonged clinical course of PCPHIV- may provide a previously unrecognized window of opportunity for clinicians to diagnose and treat patients sooner.
Our findings contrast with a retrospective review of outpatient and inpatient records of PCP patients from the 1970s and 1980s, which found that non-HIV-immunocompromised patients had a median duration of symptom onset of 5 days, vs 28 days in persons with HIV/AIDS [11]. Our discrepant findings could be because of differences in the underlying conditions of the HIV-uninfected patients in the populations studied; most patients in the previous study had hematological malignancies, whereas patients with hematological malignancies comprised less than a quarter of PCPHIV- in our study. Similarly, a prospective observational study of PCPHIV+ and PCPHIV- in 17 hospitals in France during 2007–2010 found that the median time from onset of respiratory symptoms to PCP diagnosis was significantly shorter for PCPHIV- compared with patients with PCPHIV+ (5 days vs 21 days, respectively) [19]. Again, reasons that our findings differ from this study include differences in the underlying characteristics of the non-HIV-immunocompromised patients and the fact that their study only included hospitalized patients. Our inclusion of outpatient data may provide a more complete picture of the range of severity and clinical manifestations of PCP.
In our study, most PCPHIV- had an underlying chronic pulmonary disease. Many patients with chronic pulmonary diseases are regularly treated with corticosteroid therapy, a long-recognized risk factor for PCP when given at high doses over a prolonged time period [20], and chronic pulmonary disease may itself be an underrecognized PCP risk factor [21]. A large observational study in Spain found that 15.9% of hospitalized PCP patients without HIV had a chronic pulmonary disease [22]. The increasing recognition of PCP in patients with chronic lung diseases may be partly due to increased detection by polymerase chain reaction techniques that are more sensitive than previously used diagnostic tests [23]. Distinguishing PCP infection from Pneumocystis colonization in patients with chronic pulmonary disease may be clinically challenging [9, 23]. Further research is needed on the prevention, diagnosis, and treatment of PCP in patients with chronic pulmonary diseases.
Overall, PCPHIV- had higher 90-day in-hospital mortality than PCPHIV+, but mortality rates varied among different subgroups of PCPHIV-. Whereas patients with chronic pulmonary diseases, solid organ transplants, and rheumatologic conditions had 90-day mortality similar to PCPHIV+, PCPHIV- with malignancies had a nearly 3 times higher 90-day mortality. Our study’s mortality rates in PCP patients mirror those of a previous analysis of 1999–2014 national Multiple Cause of Death data [8]. It is possible that the high mortality among PCPHIV- with malignancies is because these patients were generally older, and malignancies may involve an intrinsically higher mortality risk compared with other underlying immunosuppressive conditions. Regardless of why cancer patients with PCP experienced worse outcomes than other patients, our findings highlight the need for a high index of suspicion for PCP in immunocompromised patients with compatible symptoms.
Few patients in our study had outpatient prescriptions for PCP prophylaxis before the index date, and PCPHIV- were less likely than PCPHIV+ to have been prescribed PCP prophylaxis. Our study may in fact overestimate the number of patients prescribed PCP prophylaxis because patients can be prescribed long courses of PCP-treating drugs for conditions unrelated to Pneumocystis, and MarketScan data do not include indications for these prescriptions. Low prescribing of PCP prophylaxis in PCPHIV+ may reflect that these patients were not regularly seeking care for HIV, despite having stable health insurance coverage. For PCPHIV-, low prescribing of PCP prophylaxis may be due to the lack of comprehensive guidelines regarding prophylaxis for non-HIV-immunocompromised patients. Although published guidelines exist for PCP prophylaxis for patients with malignancy, stem cell transplants, and solid organ transplants [24–27], published guidelines do not exist for patients with chronic pulmonary diseases or rheumatologic conditions treated with immunosuppressive drugs. The American Thoracic Society suggests consideration of prophylaxis in non-HIV-immunocompromised patients who are receiving ≥20 milligrams of daily prednisone for over a month, particularly if the patient receives steroids in combination with other cytotoxic drugs or anti-TNF agents [28]. However, studies of prescribing practices among rheumatologists and infectious disease physicians reveal major inconsistencies in PCP prescribing practices [29]. Our analysis highlights the need for additional PCP prophylaxis consensus guidelines and improved adherence to existing guidelines for non-HIV-immunosuppressed persons.
Further research is needed to better understand which HIV-uninfected patients are at the greatest risk of PCP, especially because PCP prophylaxis can be difficult to tolerate and is occasionally associated with life-threatening adverse effects. A Cochrane review recommends initiating PCP prophylaxis when risk of PCP without prophylaxis is 6% [14], but PCP risk is challenging to estimate in non-HIV-immunosuppressed patients given the heterogeneity of this population. According to an analysis of the 2013 National Health Interview Survey, nearly 3% of the US population is immunosuppressed [30], and this population is poised to grow due to improved lifespans for immunocompromised patients and new indications for immunosuppressive treatments. In particular, biologic therapies for immune-mediated inflammatory diseases have been implicated as an important risk factor for invasive fungal infections [31]; the risk of PCP in patients receiving biologic therapies merits further study.
Our study has several notable limitations. The MarketScan data we used represent a large convenience sample of commercially insured patients, so our findings may not be representative of all patients at risk for PCP in the United States. Furthermore, these administrative data are subject to possible misclassification and undercoding and did not provide information regarding the PCP diagnosis method. In our study, 17% of patients had no identified underlying immunocompromising condition, and these patients had markedly lower hospitalization and mortality rates compared with other PCP patients. Though Pneumocystis can rarely cause illness in immunocompetent persons [32], we suspect that many of these patients with no underlying conditions identified were either misdiagnosed with PCP or had additional unidentified underlying conditions. Underlying conditions could have been missed if the condition was not coded by providers or because the list of ICD codes that we used to identify underlying conditions was not entirely exhaustive. Another possible source of misclassification is that some patients in this cohort may not have had incident cases of PCP if the disease was coded >3 months before the index date. A final limitation to our study is the lack of data available about drug treatments and immunosuppressive agents given in the inpatient setting. Nonetheless, our findings provide much needed insight into the current epidemiology of PCP in the United States. Our study suggests important opportunities for earlier diagnosis and treatment of PCP in HIV-uninfected patients and highlights the need for further research to define patient groups who would most benefit from PCP prophylaxis.
Supplementary Data
Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Acknowledgments
We thank Nancy Chow for early contributions to study design. We thank Zeyu Li for contributions to figure design.
Financial support. No specific funding was received for this work.
Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Potential conflicts of interest. All authors: no reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Author contributions. All authors contributed to study design, data cleaning, data analysis, and manuscript writing.
References
- 1. Edman JC, Kovacs JA, Masur H, et al. Ribosomal RNA sequence shows Pneumocystis carinii to be a member of the fungi. Nature 1988; 334:519–22. [DOI] [PubMed] [Google Scholar]
- 2. Stringer JR, Beard CB, Miller RF, Wakefield AE. A new name (Pneumocystis jiroveci) for Pneumocystis from humans. Emerg Infect Dis 2002; 8:891–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Walzer PD, Perl DP, Krogstad DJ, et al. Pneumocystis carinii pneumonia in the United States. Epidemiologic, diagnostic, and clinical features. Ann Intern Med 1974; 80:83–93. [DOI] [PubMed] [Google Scholar]
- 4. Hay JW, Osmond DH, Jacobson MA. Projecting the medical costs of AIDS and ARC in the United States. J Acquir Immune Defic Syndr (1988) 1988; 1:466–85. [PubMed] [Google Scholar]
- 5. Kaplan JE, Hanson D, Dworkin MS, et al. Epidemiology of human immunodeficiency virus-associated opportunistic infections in the United States in the era of highly active antiretroviral therapy. Clin Infect Dis 2000; 30(Suppl 1):S5–14. [DOI] [PubMed] [Google Scholar]
- 6. Buchacz K, Baker RK, Palella FJ Jr, et al. AIDS-defining opportunistic illnesses in US patients, 1994–2007: a cohort study. AIDS (London, England) 2010; 24:1549–59. [DOI] [PubMed] [Google Scholar]
- 7. Kelley CF, Checkley W, Mannino DM, Franco-Paredes C, Del Rio C, Holguin F. Trends in hospitalizations for AIDS-associated Pneumocystis jirovecii pneumonia in the United States (1986 to 2005). Chest 2009; 136:190–7. [DOI] [PubMed] [Google Scholar]
- 8. Wickramasekaran RN, Jewell MP, Sorvillo F, Kuo T. The changing trends and profile of pneumocystosis mortality in the United States, 1999–2014. 2017; 60:607–15. [DOI] [PubMed] [Google Scholar]
- 9. Morris A, Norris KA. Colonization by Pneumocystis jirovecii and its role in disease. Clin Microbiol Rev 2012; 25:297–317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kales CP, Murren JR, Torres RA, Crocco JA. Early predictors of in-hospital mortality for Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. Arch Intern Med 1987; 147:1413–7. [PubMed] [Google Scholar]
- 11. Kovacs JA, Hiemenz JW, Macher AM, et al. Pneumocystis carinii pneumonia: a comparison between patients with the acquired immunodeficiency syndrome and patients with other immunodeficiencies. Ann Intern Med 1984; 100:663–71. [DOI] [PubMed] [Google Scholar]
- 12. Monnet X, Vidal-Petiot E, Osman D, et al. Critical care management and outcome of severe Pneumocystis pneumonia in patients with and without HIV infection. Crit Care 2008; 12:R28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hughes WT, Kuhn S, Chaudhary S, et al. Successful chemoprophylaxis for Pneumocystis carinii pneumonitis. N Engl J Med 1977; 297:1419–26. [DOI] [PubMed] [Google Scholar]
- 14. Stern A, Green H, Paul M, Vidal L, Leibovici L. Prophylaxis for Pneumocystis pneumonia (PCP) in non-HIV immunocompromised patients. Cochrane Database Syst Rev 2014; 2014:CD005590. [DOI] [PubMed] [Google Scholar]
- 15. Panel on Opportunistic Infections in HIV-Infected Adults and Adolescents. Guidelines for the prevention and treatment of opportunistic infections in HIV-infected adults and adolescents:recommendations from the Centers for Disease Control and Prevention, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America Available at: http://aidsinfo.nih.gov/contentfiles/lvguidelines/oi_guidelines_pediatrics.pdf. Accessed 9 December 2019.
- 16. Benedict K, Jackson BR, Chiller T, Beer KD. Estimation of direct healthcare costs of fungal diseases in the United States. Clin Infect Dis 2019; 68:1791–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Buchacz K, Lau B, Jing Y, et al. Incidence of AIDS-defining opportunistic infections in a multicohort analysis of HIV-infected persons in the United States and Canada, 2000–2010. J Infect Dis 2016; 214:862–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Morris A, Lundgren JD, Masur H, et al. Current epidemiology of Pneumocystis pneumonia. Emerg Infect Dis 2004; 10:1713–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Roux A, Canet E, Valade S, et al. Pneumocystis jirovecii pneumonia in patients with or without AIDS, France. Emerg Infect Dis 2014; 20:1490–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Yale SH, Limper AH. Pneumocystis carinii pneumonia in patients without acquired immunodeficiency syndrome: associated illness and prior corticosteroid therapy. Mayo Clin Proc 1996; 71:5–13. [DOI] [PubMed] [Google Scholar]
- 21. Maini R, Henderson KL, Sheridan EA, et al. Increasing Pneumocystis pneumonia, England, UK, 2000–2010. Emerg Infect Dis 2013; 19:386–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Pereira-Díaz E, Moreno-Verdejo F, de la Horra C, et al. Changing trends in the epidemiology and risk factors of Pneumocystis pneumonia in Spain. Front Public Health 2019; 7:275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Fauchier T, Hasseine L, Gari-Toussaint M, et al. Detection of Pneumocystis jirovecii by quantitative PCR to differentiate colonization and pneumonia in immunocompromised HIV-positive and HIV-negative patients. J Clin Microbiol 2016; 54:1487–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Tomblyn M, Chiller T, Einsele H, et al. ; Center for International Blood and Marrow Research ; National Marrow Donor program; European Blood and MarrowTransplant Group; American Society of Blood and Marrow Transplantation; Canadian Blood and Marrow Transplant Group; Infectious Diseases Society of America; Society for Healthcare Epidemiology of America; Association of Medical Microbiology and Infectious Disease Canada; Centers for Disease Control and Prevention. Guidelines for preventing infectious complications among hematopoietic cell transplantation recipients: a global perspective. Biol Blood Marrow Transplant 2009; 15:1143–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Baden LR, Swaminathan S, Angarone M, et al. Prevention and treatment of cancer-related infections, version 2.2016, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw 2016; 14:882–913. [DOI] [PubMed] [Google Scholar]
- 26. Cooley L, Dendle C, Wolf J, et al. Consensus guidelines for diagnosis, prophylaxis and management of Pneumocystis jirovecii pneumonia in patients with haematological and solid malignancies, 2014. Intern Med J 2014; 44:1350–63. [DOI] [PubMed] [Google Scholar]
- 27. Martin SI, Fishman JA; AST Infectious Diseases Community of Practice Pneumocystis pneumonia in solid organ transplantation. Am J Transplant 2013; 13(Suppl 4):272–9. [DOI] [PubMed] [Google Scholar]
- 28. Limper AH, Knox KS, Sarosi GA, et al. ; American Thoracic Society Fungal Working Group An official American Thoracic Society statement: treatment of fungal infections in adult pulmonary and critical care patients. Am J Respir Crit Care Med 2011; 183:96–128. [DOI] [PubMed] [Google Scholar]
- 29. Wolfe RM, Beekmann SE, Polgreen PM, Winthrop KL, Peacock JE Jr. Practice patterns of Pneumocystis pneumonia prophylaxis in connective tissue diseases: a survey of infectious disease physicians. Open Forum Infect Dis 2019; 6:XXX–XX. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Harpaz R, Dahl RM, Dooling KL. Prevalence of immunosuppression among US adults, 2013. JAMA 2016; 316:2547–8. [DOI] [PubMed] [Google Scholar]
- 31. Vallabhaneni S, Chiller TM. Fungal infections and new biologic therapies. Curr Rheumatol Rep 2016; 18:29. [DOI] [PubMed] [Google Scholar]
- 32. Cano S, Capote F, Pereira A, et al. Pneumocystis carinii pneumonia in patients without predisposing illnesses. Acute episode and follow-up of five cases. Chest 1993; 104:376–81. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.