Abstract
Laboratory testing is required to distinguish coccidioidomycosis and histoplasmosis from other types of community-acquired pneumonia (CAP). In this nationwide survey of 1258 health care providers, only 3.7% reported frequently testing CAP patients for coccidioidomycosis and 2.8% for histoplasmosis. These diseases are likely underdiagnosed, and increased awareness is needed.
Keywords: coccidioidomycosis, histoplasmosis, pneumonia, primary health care, United States
Inhalation of the fungal pathogens Coccidioides and Histoplasma can result in a wide range of illnesses, ranging from subclinical infection to life-threatening disseminated disease. Community-acquired pneumonia (CAP) is a common illness caused by various pathogens [1, 2] and is a frequent presentation for coccidioidomycosis and histoplasmosis. Because the clinical features of coccidioidomycosis and histoplasmosis can resemble those of other pneumonias, these diseases are often misdiagnosed and inappropriately treated with antibacterial medications [3]. Many coccidioidomycosis and histoplasmosis patients first present to primary care providers (PCPs) but experience delays of several weeks before being tested by infectious disease or pulmonary specialists [4, 5]. Such diagnostic delays contribute to increased health care costs and poorer patient outcomes [6, 7].
In the United States, coccidioidomycosis is most common in Arizona and California and histoplasmosis in the Midwest and South, but both diseases have broad endemic regions, and travel-associated cases occur regularly. Previous health care provider (HCP) surveys have examined coccidioidomycosis testing practices in Arizona and Washington State [8, 9], but the topic has not been assessed nationwide, and no similar data exist for histoplasmosis. To inform strategies to improve HCP awareness of coccidioidomycosis and histoplasmosis, we assessed self-reported testing frequency and HCP-related features associated with testing for these diseases among CAP patients.
METHODS
We used data from the Spring 2020 DocStyles survey, a web-based survey of primary care physicians, obstetricians and gynecologists (OB/GYNs), pediatricians, nurse practitioners, and physician assistants. The survey was commissioned by Porter Novelli Public Services, a public relations firm, and conducted by SERMO, a global market research company. Respondents were screened to include only HCPs actively seeing patients in the United States and practicing for at least 3 years. Porter Novelli developed the survey instrument with guidance from federal public health agencies and other clients. The survey aimed to evaluate HCPs’ attitudes and practices for various health conditions and to assess their use of medical information and continuing education sources. We asked 2 questions about HCPs’ (excluding OB/GYNs and pediatricians) testing practices: “How often do you test for [coccidioidomycosis (1) or histoplasmosis (2)] in patients with community-acquired pneumonia (CAP)?” Response options were “frequently,” “sometimes,” “rarely,” “never,” and “I do not see patients with CAP.”
We performed descriptive analyses and Type 3 likelihood ratio tests to select demographic features and information-seeking behaviors associated with testing frequency. Multivariable ordered logistic regression was used to estimate adjusted odds ratios (aORs) and 95% confidence intervals. Analyses were conducted using SAS (version 9.4; SAS Institute). Because no personally identifiable information was used during this analysis of data collected by Porter Novelli, the analysis was not subject to review by CDC’s institutional review board.
RESULTS
Of 2620 HCPs surveyed, 1760 (67.2%) completed the survey, and 1258 were asked the fungal disease testing questions. Of the 1258, most were internists (45.2%) or family practice physicians (34.8%), and 83.0% worked primarily in outpatient settings (PCPs). The mean number of years in practice was 15.3.
Overall, 3.7% reported “frequently” and 15.0% reported “sometimes” testing CAP patients for coccidioidomycosis (Table 1). Among 43 HCPs in Arizona, 32.4% reported “frequently” and 38.2% reported “sometimes” testing for coccidioidomycosis; among 148 HCPs in California, 7.4% reported “frequently” testing, and 29.7% reported “sometimes” testing. Nationwide, 2.8% of HCPs reported “frequently” testing for histoplasmosis, and 19.0% reported “sometimes” testing.
Table 1.
Health Care Providers’ Self-Reported Testing Frequency for Coccidioidomycosis and Histoplasmosis Among Community-Acquired Pneumonia Patients, United States, 2020
| Coccidioidomycosis, No. (%) | Histoplasmosis, No. (%) | |
|---|---|---|
| Frequently | 47 (3.7) | 35 (2.8) |
| Sometimes | 189 (15.0) | 239 (19.0) |
| Rarely | 488 (38.8) | 527 (41.9) |
| Never | 435 (34.6) | 364 (28.9) |
| I do not see patients with CAP | 99 (7.9) | 93 (7.4) |
Abbreviation: CAP, community-acquired pneumonia.
For coccidioidomycosis, the odds of more frequent testing were 36% higher among internists compared with family practice providers (aOR, 1.36; 95% CI, 1.04–1.77) and were 72% higher among HCPs who identified as “other” race compared with White (aOR, 1.72; 95% CI, 1.12–2.65) (Table 2). Higher odds of more frequent testing for coccidioidomycosis and histoplasmosis occurred among HCPs in the West (aOR, 5.17; 95% CI, 3.64–7.35; aOR, 1.53; 95% CI, 1.09–2.15) and the Midwest (aOR, 1.61; 95% CI, 1.15–2.25; aOR, 1.69; 95% CI, 1.22–2.34) compared with the Northeast. Higher odds of more frequent testing also occurred among HCPs with teaching hospital privileges (aOR, 1.54; 95% CI, 1.20–1.98; aOR, 1.49; 95% CI, 1.17–1.89) and those who primarily worked in an inpatient setting (aOR, 1.52; 95% CI, 1.09–2.11; aOR, 2.03; 95% CI, 1.46–2.81), compared with a group outpatient practice setting.
Table 2.
Health Care Provider–Related Factors Associated With Frequently Testing Community-Acquired Pneumonia Patients for Coccidioidomycosis and Histoplasmosis, United States, 2020
| Coccidioidomycosis | Histoplasmosis | ||||||
|---|---|---|---|---|---|---|---|
| No. (%) (total n = 1258) | aOR | 95% CI | Type 3 χ 2 P Valuea | aOR | 95% CI | Type 3 χ 2 P Valuea | |
| Demographic characteristics | |||||||
| Mean age (SD), y | 45 (11.1) | 1.01 | 0.99–1.02 | .4050 | 1.00 | 0.99–1.01 | .6892 |
| Female | 455 (36.2) | 0.88 | 0.68–1.15 | .3503 | 0.98 | 0.76–1.27 | .8701 |
| Hispanic | 66 (5.2) | 1.10 | 0.67–1.80 | .7033 | 1.28 | 0.78–2.08 | .3272 |
| Race | .0239 | .4324 | |||||
| White | 862 (68.5) | Ref | Ref | Ref | Ref | ||
| Black | 38 (3.0) | 0.62 | 0.31–1.23 | 0.65 | 0.33–1.25 | ||
| Asian | 268 (21.3) | 1.22 | 0.92–1.61 | 1.02 | 0.77–1.35 | ||
| Otherb | 90 (7.2) | 1.72 | 1.12–2.65 | 1.23 | 0.80–1.90 | ||
| Region | <.0001 | .0116 | |||||
| Northeast | 281 (22.3) | Ref | Ref | Ref | Ref | ||
| Midwest | 292 (23.2) | 1.61 | 1.15–2.25 | 1.69 | 1.22–2.34 | ||
| South | 427 (33.9) | 1.34 | 0.97–1.83 | 1.32 | 0.97–1.79 | ||
| West | 258 (20.5) | 5.17 | 3.64–7.35 | 1.53 | 1.09–2.15 | ||
| Practice characteristics | |||||||
| Provider specialty | .0052 | .0511 | |||||
| Family practice | 438 (34.8) | Ref | Ref | Ref | Ref | ||
| Internist | 569 (45.2) | 1.36 | 1.04–1.77 | 1.25 | 0.96–1.63 | ||
| Nurse practitioner | 129 (10.3) | 0.73 | 0.47–1.13 | 1.02 | 0.67–1.55 | ||
| Physician assistant | 122 (9.7) | 0.72 | 0.46–1.12 | 0.70 | 0.45–1.08 | ||
| Patients per week | 106.9 (70.6) | 1.00 | 1.00–1.01 | .0002 | 1.00 | 1.00-1.00 | |
| Teaching hospital privileges | 567 (45.1) | 1.54 | 1.20–1.98 | .0006 | 1.49 | 1.17–1.89 | .0014 |
| Practice setting | .0299 | <.0001 | |||||
| Group outpatient practice | 844 (67.1) | Ref | Ref | Ref | Ref | ||
| Individual outpatient practice | 201 (15.9) | 1.23 | 0.90–1.68 | 1.06 | 0.78–1.44 | ||
| Inpatient practice | 213 (16.9) | 1.52 | 1.09–2.11 | 2.03 | 1.46–2.81 | ||
| Medical news sources, mean frequency (SD)c | |||||||
| Books | 2.8 (1.0) | 1.38 | 1.22–1.55 | <.0001 | 1.42 | 1.26–1.59 | <.0001 |
| Government health agencies | 3.5 (1.0) | 1.05 | 0.92–1.21 | .4793 | 1.05 | 0.92–1.21 | .4684 |
| Pharmaceutical companies/representatives | 2.9 (1.1) | 1.27 | 1.14–1.42 | <.0001 | 1.32 | 1.18–1.47 | <.0001 |
| Medical websites | 4.0 (0.9) | 0.97 | 0.84–1.12 | .6971 | 0.96 | 0.84–1.11 | .5819 |
| Magazine stories/articles | 2.9 (1.0) | 1.02 | 0.90–1.15 | .8029 | 0.91 | 0.81–1.03 | .1448 |
| Medical journals | 3.9 (0.9) | 1.14 | 0.98–1.32 | .0806 | 1.19 | 1.03–1.37 | .0189 |
| Newspaper stories/articles | 2.9 (1.0) | 1.08 | 0.95–1.23 | .2583 | 1.11 | 0.98–1.26 | .1031 |
| Physicians | 4.0 (0.9) | 0.96 | 0.83–1.1 | .5303 | 1.08 | 0.94–1.24 | .2900 |
| Professional medical societies | 3.5 (1.0) | 1.18 | 1.04–1.35 | .0130 | 1.18 | 1.03–1.34 | .0137 |
| Search engines | 3.8 (1.0) | 0.84 | 0.74–0.96 | .0073 | 0.85 | 0.75–0.96 | .0078 |
| Social media | 3.0 (1.2) | 1.09 | 0.98–1.22 | .1297 | 1.07 | 0.96–1.20 | .2115 |
| Mobile applications | 3.6 (1.1) | 1.12 | 1.00–1.26 | .0517 | 1.09 | 0.97–1.21 | .1566 |
| Continuing medical education sourcesd | |||||||
| Internet sites | 813 (64.6) | 0.78 | 0.61–0.99 | .0435 | 0.92 | 0.73–1.17 | .5152 |
| Conferences | 862 (68.5) | 1.11 | 0.87–1.43 | .4035 | 0.98 | 0.77–1.26 | .8907 |
| Journals | 860 (68.4) | 0.67 | 0.52–0.86 | .0020 | 0.75 | 0.58–0.97 | .0261 |
| Government health agencies | 531 (42.2) | 1.21 | 0.95–1.55 | .1283 | 1.12 | 0.88–1.42 | .3606 |
| Classes | 372 (29.6) | 1.13 | 0.88–1.44 | .3460 | 0.95 | 0.75–1.21 | .6828 |
| CD-ROM | 58 (4.6) | 2.13 | 1.28–3.55 | .0038 | 3.77 | 2.24–6.32 | <.0001 |
| Medical podcasts | 460 (36.6) | 1.04 | 0.82–1.32 | .7538 | 1.13 | 0.89–1.43 | .3078 |
| Something else not listed | 83 (6.6) | 0.89 | 0.57–1.4 | .6088 | 0.94 | 0.61–1.46 | .7949 |
| No CME in the past year | 16 (1.3) | 0.35 | 0.12–1.09 | .0692 | 0.88 | 0.31–2.45 | .7993 |
| Likelihood ratio test for global beta = 0e | 354.43 | 298.68 | |||||
aORs correspond to ordered logistic regression for the answer choices “frequently,” “sometimes,” “rarely,” and “never/I do not see patients with CAP” and represent the odds of more frequent testing associated with a 1-unit change in the explanatory variable, controlling for all other explanatory variables. Additional factors evaluated in bivariate analyses but not included in the final models include years practicing medicine, number of providers in group practice, telemedicine use, social media use, email use for communication with patients, and approximate household income of most patients. The ordered logistic regression assumes proportional odds, that is, that the slopes between testing frequencies are constant across the range of testing frequencies. This assumption is violated here, but comparison with multinomial results did not reveal differences in slope signs or substantial differences in magnitude.
Abbreviations: aOR, adjusted odds ratio; CAP, community-acquired pneumonia; CME, continuing medical education.
aType 3 likelihood ratio test, used to select variables for inclusion in the analyses, evaluates overall contribution of each explanatory variable to the model fit. The null hypothesis is that the marginal contribution is 0.
b“Other” = “Native Hawaiian or other Pacific Islander,” “American Indian or Alaska Native,” “2 or more races,” or “other race.”
cMean and standard deviation correspond to values for the question “How often do you use each of the following to keep up to date with the latest medical news and trends?” with answer choices “never = 1,” “rarely = 2,” “sometimes = 3,” “often = 4,” and “regularly = 5.”
dPercentage of participants who responded “yes” to each source for the question “Which of the following sources have you used to pursue continuing medical education in the past year?”
eThe likelihood ratio test with 37 degrees of freedom tests the global null hypothesis that global BETA = 0. Reject for both ordered logistic regressions with P < .0001.
Sources of medical information most strongly associated with more frequent testing for coccidioidomycosis and histoplasmosis included books (aOR, 1.38; 95% CI, 1.22–1.55; aOR, 1.42; 95% CI, 1.26–1.59), pharmaceutical companies and representatives (aOR, 1.27; 95% CI, 1.14–1.42; aOR, 1.32; 95% CI, 1.18–1.47), and medical journals (aOR, 1.14; 95% CI, 0.98–1.32; aOR, 1.19; 95% CI, 1.03–1.37). Not having participated in any continuing medical education activities in the past year was not associated with testing frequency.
DISCUSSION
Testing for coccidioidomycosis and histoplasmosis among patients with CAP was low in this nationwide survey of primarily PCPs, adding to evidence that coccidioidomycosis and histoplasmosis are likely widely underdiagnosed. Factors associated with testing frequency appeared to reflect the patient population rather than provider-related features. Overall, these results indicate that continued efforts to increase awareness of and testing for coccidioidomycosis and histoplasmosis are needed, particularly in the primary care setting, to help facilitate earlier diagnosis and treatment.
The low testing rates we observed might indicate that many HCPs are unaware of the importance of testing for coccidioidomycosis and histoplasmosis, which are frequently misdiagnosed and often cause prolonged illness [4, 5]. The American Thoracic Society (ATS)/Infectious Diseases Society of America (IDSA) guidelines for diagnosing CAP in adults do not explicitly recommend testing for endemic fungi, stating that they are uncommon pathogens in CAP [1]. The contribution of Coccidioides and Histoplasma relative to other CAP etiologies nationwide is not well established (a landmark study found no bacterial or viral pathogen in 62% of hospitalized CAP patients, raising the possibility that some unidentified pneumonias could be caused by fungal pathogens) [10]. However, the numbers of reported coccidioidomycosis cases (despite being reportable in only half of all states) are similar to legionellosis [11], for which the ATS/IDSA guidelines recommend testing in certain epidemiological circumstances and among patients with severe CAP. Newly developed ATS guidelines for diagnosing fungal infections provide important direction about serology and antigen testing, but the focus on pulmonary and critical care practice leaves a gap in guidance for other practice settings, potentially resulting in missed opportunities for earlier diagnosis [12].
The differences in testing by geographic region we observed are consistent with foci of histoplasmosis in the Midwest and coccidioidomycosis in the West, with higher testing rates in Arizona and California, the 2 states with the highest geographic risk for coccidioidomycosis. Similar to our results, a 2007 survey of Arizona HCPs found that 74% reported testing CAP patients for coccidioidomycosis at least 50% of the time and that testing frequency increased in highly endemic areas [8]. Although testing for coccidioidomycosis and histoplasmosis may not be warranted in all CAP patients, increased suspicion in areas where they are most common is warranted, along with periodic reevaluation of how these areas may have changed [13].
Factors influencing testing decisions are likely both patient- and provider-related. Previous analyses have identified CAP patient race/ethnicity and other demographic and clinical features associated with being tested for coccidioidomycosis [14]. This study adds insight into the role of provider features; aside from race, which may reflect providers’ awareness of their own risk [15] for coccidioidomycosis, personal features did not appear to influence testing frequency. Instead, higher odds of more frequent testing among internists, those with teaching hospital privileges, and those who work in the inpatient setting appear to reflect HCPs who care for more seriously ill patients. They may treat patients who initially present with more severe disease or those whose illness progressed after delayed diagnosis, although our study was not able to assess timing of fungal disease testing, and the data are limited to self-report rather than actual testing practices. Social desirability bias would suggest that testing occurs even less often than reported. This survey also did not collect information about test types or HCPs’ confidence in their ability to diagnose fungal diseases, which would be useful for a deeper understanding of testing practices. Future surveys could also evaluate testing for blastomycosis, another geographically focused disease that can cause CAP, as well as target HCPs who work in emergency medicine, another setting where coccidioidomycosis and histoplasmosis patients commonly first present for care [4, 5].
Nationwide, HCPs’ testing for coccidioidomycosis and histoplasmosis among CAP patients was low. More studies are needed to establish the true burden of these fungal diseases and fungal pneumonia compared with other causes of CAP, and CAP guideline developers could consider providing more specific guidance about fungal disease testing.
Acknowledgments
We thank Fred Fridinger, DrPH, in the CDC Office of the Associate Director for Communication, and Deanne Weber, PhD, Porter Novelli Public Services, Inc., for coordinating access to the DocStyles data.
Potential conflicts of interest. All authors declare no 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.
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.
Participant consent. This study does not include factors necessitating participant consent.
References
- 1. Metlay JP, Waterer GW, Long AC, et al. Diagnosis and treatment of adults with community-acquired pneumonia. An official clinical practice guideline of the American Thoracic Society and Infectious Diseases Society of America. Am J Respir Crit Care Med 2019; 200:e45–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Musher DM, Thorner AR. Community-acquired pneumonia. N Engl J Med 2014; 371:1619–28. [DOI] [PubMed] [Google Scholar]
- 3. Hage CA, Knox KS, Wheat LJ. Endemic mycoses: overlooked causes of community acquired pneumonia. Respir Med 2012; 106:769–76. [DOI] [PubMed] [Google Scholar]
- 4. Tsang CA, Anderson SM, Imholte SB, et al. Enhanced surveillance of coccidioidomycosis, Arizona, USA, 2007-2008. Emerg Infect Dis 2010; 16:1738–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Benedict K, McCracken S, Signs K, et al. Enhanced surveillance for histoplasmosis—nine states, 2018–2019. Open Forum Infect Dis 2020; 7:ofaa343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Ginn R, Mohty R, Bollmann K, et al. Delays in coccidioidomycosis diagnosis and relationship to healthcare utilization, Phoenix, Arizona, USA. Emerg Infect Dis 2019; 25:1742–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hage CA, Bowyer S, Tarvin SE, et al. Recognition, diagnosis, and treatment of histoplasmosis complicating tumor necrosis factor blocker therapy. Clin Infect Dis 2010; 50:85–92. [DOI] [PubMed] [Google Scholar]
- 8. Chen S, Erhart LM, Anderson S, et al. Coccidioidomycosis: knowledge, attitudes, and practices among healthcare providers—Arizona, 2007. Med Mycol 2011; 49:649–56. [DOI] [PubMed] [Google Scholar]
- 9. Hubber J, Person A, Jecha L, et al. Knowledge, attitudes, and practices regarding coccidioidomycosis among healthcare providers in four counties in Washington State, 2017. Med Mycol 2020; 58:730–6. [DOI] [PubMed] [Google Scholar]
- 10. Jain S, Self WH, Wunderink RG, et al. ; CDC EPIC Study Team Community-acquired pneumonia requiring hospitalization among U.S. adults. N Engl J Med 2015; 373:415–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. National Notifiable Diseases Surveillance System (NNDSS) Available at: https://wwwn.cdc.gov/nndss/data-and-statistics.html. Accessed 13 August 2020.
- 12. Hage CA, Carmona EM, Epelbaum O, et al. Microbiological laboratory testing in the diagnosis of fungal infections in pulmonary and critical care practice. An official American Thoracic Society clinical practice guideline. Am J Respir Crit Care Med 2019; 200:535–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ashraf N, Kubat RC, Poplin V, et al. Re-drawing the maps for endemic mycoses. Mycopathologia 2020; 185:843–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Tartof SY, Benedict K, Xie F, et al. Testing for coccidioidomycosis among community-acquired pneumonia patients, Southern California, USA. Emerg Infect Dis 2018; 24:779–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Berger JT The influence of physicians’ demographic characteristics and their patients’ demographic characteristics on physician practice: implications for education and research. Acad Med 2008; 83:100–5. [DOI] [PubMed] [Google Scholar]