OBJECTIVE: This study explores the use of spirometry in primary care settings.
METHODS: A 4-page survey was mailed to a national, random sample of office-based family physicians and pediatricians. Survey items addressed knowledge, attitudes, and practices regarding spirometry and standardized clinical vignettes. Data were analyzed by using χ2 tests and multivariate logistic regression.
RESULTS: Among the 360 respondents who provided care to children with asthma, 52% used spirometry in clinical practice, whereas 80% used peak flow meters and 10% used no lung function tests. Only 21% routinely used spirometry for all guideline-recommended clinical situations. More family physicians than pediatricians reported using spirometry (75% vs 35%; P < .0001), and family physicians were more comfortable in interpreting spirometric results (50% vs 25%; P < .0001). Only one-half of respondents interpreted correctly the spirometric results in a standardized clinical vignette, and the frequency of underrating asthma severity increased with the inclusion of spirometric results. The most common barriers to the use of spirometry, that is, time and training, were cited more often by physicians who did not use spirometry. Two-thirds of respondents agreed that they would want additional training regarding implementing spirometry in their clinical practices.
CONCLUSIONS: The use of spirometry in primary care settings for children with asthma does not conform to national guidelines. Widespread implementation of national asthma guidelines likely would require a major educational initiative to address deficiencies in spirometry interpretation and other barriers.
WHAT'S KNOWN ON THIS SUBJECT:
National guidelines consider spirometry testing to be a key clinical activity for the diagnosis and management of asthma, but spirometry use by pediatric primary care physicians for asthma is thought to be limited.
WHAT THIS STUDY ADDS:
Important shortcomings in spirometry testing are commonplace among pediatric primary care providers, especially among pediatricians. Widespread implementation of national asthma management guidelines likely would require major spirometry training initiatives among pediatric primary care providers.
The National Asthma Education and Prevention Program (NAEPP) Expert Panel Report guidelines underscore the importance of spirometry in the initial diagnosis of asthma and at least once per year thereafter for ongoing management.1 Spirometry can be performed reliably for children as young as 6 years of age, although it requires a spirometer, personnel trained to administer the test under the American Thoracic Society guidelines,2 and physicians with knowledge to interpret the results. Peak flow is an alternative measure of lung function that can be monitored easily with an inexpensive device, although results recorded during a single encounter may be highly variable. Consequently, peak flow assessment is not considered a reliable method for the diagnosis of asthma3,–,5 and is not endorsed by the NAEPP for asthma diagnosis. In contrast, spirometry provides measures of lung volumes and airflow during a forced expiratory maneuver, with a high degree of reproducibility among patients. Previous studies suggested that physicians often underestimate asthma severity in the absence of spirometric results6,–,8 and the course of treatment may be altered substantially when spirometric information is made available.6
Despite longstanding NAEPP guidelines, primary care physicians' use of spirometry for asthma management among children remains limited.9,10 However, it is not clear what practice-level barriers may hinder the use of spirometry by primary care physicians. The objectives of this study were (1) to describe the use of spirometry among physicians who provide primary care to children with asthma, (2) to assess physician knowledge, attitudes, and preferences regarding the use of spirometry, and (3) to assess the extent to which perceived barriers to the use of spirometry exist among pediatric primary care providers.
We conducted a mail survey of office-based general pediatricians and family physicians, focusing on knowledge, attitudes, and practices regarding and perceived barriers to the use of spirometry. The study was approved by the institutional review board of the University of Michigan. We obtained a national, random sample of 500 family physicians and 500 general pediatricians from the American Medical Association Masterfile through a contracted vendor (Medical Marketing Services, Wood Dale, IL). The Masterfile is the most-comprehensive listing of licensed physicians in the United States and includes both American Medical Association members and nonmembers. The sampling frame included all allopathic and osteopathic physicians providing direct, office-based, patient care; we excluded physicians ≥70 years of age, resident physicians, and physicians practicing at federal (eg, Veterans Administration) facilities.
The investigators developed and refined survey items that reflected the primary study aims, drawn from the published literature on spirometry, clinical practice experience, and the NAEPP asthma guidelines. Survey items included physicians' use of spirometry and peak flow meters in clinical practice, physicians' attitudes toward lung function testing, perceived barriers to the use of spirometry, and demographic information (Appendix 1). The survey also included a clinical vignette to assess the degree to which spirometry may affect physicians' recognition of asthma severity and subsequent medication choices. The vignette first described a clinical history for a pediatric patient with mild persistent asthma. After assessment of this interpretation, the vignette continued by providing spirometric results for the case patient that described a moderate obstruction, which elevated the patient's asthma severity to moderate persistent (Fig 1). Survey questions encompassed a variety of formats, including Likert scales and fixed-choice response items. Questions were pilot-tested with a convenience sample of physicians to assess clarity and ease of administration; revisions were made on the basis of their feedback. The survey mailings included a personalized cover letter inviting participation and a postage-paid reply envelope. The initial mailing was completed in December 2007 and included a $2 cash incentive. Two subsequent mailings were sent to nonrespondents at 4-week intervals.
Response frequencies were assessed for differences between groups on the basis of specialty and spirometry user status. Spirometry users were defined as providers who reported that they use spirometry in clinical practice with adults or children. Multivariate logistic regression analysis was conducted to assess associations between physician characteristics and the use of spirometry by using adjusted odds ratios (ORs) with 95% confidence intervals (CIs). Likelihood χ2 tests were used to detect bivariate associations between the use of spirometry, physicians' specialty, and physicians' knowledge and attitudes regarding spirometry; P values of <.05 were considered statistically significant. Clinical vignette analyses were conducted according to spirometry user status and specialty. All analyses were conducted by using SAS 9.1 (SAS Institute, Cary, NC).
Of the 1000 physicians in the mailing sample, 50 (5%) were excluded because the mailed materials were returned as undeliverable. Completed surveys were returned by 477 respondents, for an overall response rate of 50% (family physicians, 47%; pediatricians, 54%). The geographic distribution of respondents (Northeast, 17%; Midwest, 25%; South, 35%; West, 23%) was similar to that of nonrespondents (P = .0932). Among the 477 respondents, 117 (25%) indicated that they did not provide outpatient care to children with asthma (reported primarily by family physicians) and were excluded from analyses. The remaining 360 responses eligible for analyses represented 150 family physicians and 210 pediatricians. Eligible respondents were more likely than survey nonrespondents (N = 473) to be board-certified (91% vs 85%; P = .005); there were no significant differences between respondents and nonrespondents in terms of gender, degree type, or years since graduation from medical school.
Use of Spirometry
Overall, 52% of respondents indicated that they used spirometry in clinical practice, and use was more common among family physicians than among pediatricians (75% vs 35%; P < .0001); 51% of family physicians reported that they used spirometry in the treatment of both adults and children. Multivariate logistic regression analysis confirmed more-frequent use among family physicians, compared with pediatricians (OR: 7.6 [95% CI: 3.7–15.4]), with controlling for other characteristics (Table 1). Similarly, the use of spirometry was reported more frequently by osteopathic physicians than by allopathic physicians (OR: 4.7 [95% CI: 1.6–13.4]) and by physicians who saw >15 pediatric asthma patients per week than by physicians who saw <5 patients per week (OR: 2.6 [95% CI: 1.0–6.8]). The use of spirometry was not associated with board certification or practice affiliation. Among physicians who used spirometry for adults and children (N = 147), approximately one-third indicated routine use in clinical situations as recommended by national guidelines: establishing an asthma diagnosis, 36%; classifying asthma severity, 31%; classifying asthma control, 34%. Only 21% reported routine use of spirometry for all 3 recommended clinical situations. Peak flow meter use was reported more frequently (80%) than was the use of spirometry and also was more common among family physicians than among pediatricians (87% vs 76%; P = .009). Importantly, 10% of respondents reported using no lung function tests in their practice, which varied sharply between family physicians (3%) and pediatricians (15%; P < .0001).
Attitudes Regarding Lung Function Testing
Physicians' reported use of spirometry was not necessarily aligned with their attitudes and preferences regarding lung function testing. Only 35% of physicians agreed or strongly agreed that they were comfortable interpreting spirometric results; family physicians reported higher comfort levels, compared with pediatricians (50% vs 25%; P < .0001). The rate of spirometry use was higher among physicians who reported being comfortable interpreting spirometric results, compared with those who were not comfortable (83% vs 22%; P < .0001). In contrast, 79% of respondents agreed or strongly agreed that they were comfortable interpreting peak flow meter results, with no difference between spirometry users and nonusers or according to specialty. Most respondents (73%) agreed or strongly agreed that “spirometry is more reliable than peak flow for classifying the severity of asthma,” with a higher rate of agreement among pediatricians than among family physicians (80% vs 64%; P = .003). In contrast, only 13% of respondents agreed or strongly agreed that “peak flow was as good as spirometry for the diagnosis of asthma.” Spirometry users exhibited more positive attitudes than nonusers regarding implementation of spirometry in primary care practice settings (Fig 2). More spirometry users than nonusers agreed that there was sufficient time during appointments to interpret spirometric results and that their staff members were trained adequately to perform spirometry. Two-thirds of respondents agreed or strongly agreed that they would want additional training regarding implementation of spirometry in their clinical practices. This view was more common among pediatricians than among family physicians (78% vs 53%; P < .0001). Respondents indicated that their preferred methods for spirometry training included workshops with asthma care specialists (43%), continuing medical education seminars (32%), Internet-based training (24%), and learning from colleagues (14%). No single method was endorsed by more than one-half of respondents, and there were no differences in preferred training methods according to specialty.
Interpretation of Spirometry Results
The spirometric results from the clinical vignette were interpreted correctly as a moderate obstruction by 49% of respondents; spirometry users (54%) were more likely than nonusers (47%) to select the correct interpretation (P < .0001). Notably, 14% of respondents reported that they did not know how to interpret the provided spirometric results. When they were asked to apply the spirometric results to severity assessment for the patient, 65% of respondents rated the asthma severity correctly as moderate persistent, with no difference between spirometry users and nonusers (Fig 1). Of those who identified the mild persistent status correctly on the basis of the clinical history alone, approximately one-half (54%) increased the severity classification correctly to moderate persistent after reading the spirometric results. Importantly, the frequency of underrated severity increased substantially with the inclusion of spirometric results (from 15% to 33%); two-thirds of physicians who underrated the severity on the basis of the clinical history in conjunction with spirometric results had rated the severity correctly on the basis of symptoms alone. Assessment of asthma severity, either on the basis of clinical history alone or with the addition of spirometric results, did not differ according to specialty or spirometry use. The majority of physicians (87%) correctly chose inhaled corticosteroids and albuterol as the appropriate medication regimen, with no difference between spirometry users and nonusers. Even among respondents who underrated asthma severity when given both symptoms and spirometric results, 76% chose the correct medication regimen.
In contrast to NAEPP guidelines, our findings indicate that pediatric primary care physicians use spirometry in their clinical practices to a limited degree and use is far less common among pediatricians than among family physicians. Current national asthma guidelines emphasize spirometric testing for the diagnosis of asthma because clinical history and physical examination findings alone are not reliable for this purpose.1 Importantly, delayed diagnoses of asthma or inaccurate assessment of severity can lead to permanently decreased lung function11,12 and elevated risk of hospitalization or death.13,–,15 The finding that few pediatricians use spirometry is consistent with previous reports of widespread underuse and unfamiliarity with spirometric testing in primary care settings10,16,–,18 and for pediatric asthma.9 Instead, we found peak flow meters to be widely used in clinical practice. Peak flow measures are not recommended for diagnostic purposes because readings from the devices can be highly variable, even within the same encounter, and appropriate reference values for predicted expiratory flow typically are not available.1
Spirometric testing was shown previously to be feasible in primary care settings and to provide valuable details for the diagnosis and management of respiratory disease,17,19 even among children as young as 6 years of age.20 However, we found several physician perceptions that may hinder more-widespread use of spirometry, notably with respect to adequate training of staff members for spirometric testing, sufficient time, or adequate reimbursement for spirometry; these concerns were cited more frequently by physicians who did not use spirometry.
Importantly, most pediatric primary care physicians reported that they were not comfortable interpreting spirometric results. Unfamiliarity or inexperience with spirometry might contribute to these perceptions, which might be allayed through adequate training of physicians and staff members. Given the findings of this study, it is likely that primary care pediatricians have little exposure to spirometry during their training. Ensuring that pediatric and family medicine residents receive training in interpretation of spirometric findings during their pulmonology rotation would be beneficial, as would the development of simulation exercises and computerized testing modules focused on improving the understanding and interpretation of spirometric findings. Concerns regarding lack of time may reflect primary care physicians' attitudes regarding reimbursement for spirometric testing. We found that few primary care physicians thought that spirometric testing was reimbursed adequately, which was cited previously as a barrier in other settings.10,17,21,22 Previous studies found that pay-for-performance programs can be an effective mechanism to influence physician behavior in several areas of patient care,23 including asthma management,24 although the effectiveness of pay-for-performance programs addressing specifically the use of spirometry is not known.
Physicians were only modestly successful in the interpretation of spirometric results and the subsequent recognition of asthma. We found deficiencies in physician recognition of asthma severity both with and without spirometric results, regardless of whether physicians reported using spirometry in their regular clinical practices. Although spirometric results improved the overall accuracy of severity assessment by physicians, there was a troubling shift toward underrating asthma severity, which suggests that important gaps exist with respect to accurate application of spirometric results. These findings contradict previous studies that found that, in the absence of objective information from spirometry, physicians typically underestimate the severity of asthma.6,–,8 One study found that, when spirometry information was made available, it substantially altered the course of treatment.6 In that study, however, the spirometric results were interpreted primarily by pulmonologists and not primary care providers.
Our findings, particularly those related to deficiencies in primary care physicians' comfort level and ability to interpret spirometric results, suggest that additional physician training will be necessary to realize more fully the benefits of spirometric testing. Numerous spirometry training programs are currently available, using face-to-face, computer-based, and Internet-based teaching modalities; 32 US training programs have been approved by the National Institute for Occupational Safety and Health (NIOSH).25 However, the NIOSH–approved training programs fulfill a prescribed purpose in the context of occupational health and therefore are not likely to be widely used by primary care providers. In addition, a practical guide for spirometric testing in primary care settings has been published.26 These training opportunities may need greater promotion among primary care physicians, particularly among general pediatricians. Importantly, evidence suggests that spirometry training on an ongoing basis may be needed.27
There are noteworthy limitations to this study. First, although the response rate of 50% is comparable to rates for other physician surveys,17 our results may not be representative of the population of pediatric primary care providers. This survey assessed knowledge, attitudes, and self-reported behaviors regarding the use of spirometry; independent verification of these measures was not available. In addition, we did not assess the sites where spirometry was performed, the specific equipment used, the types of staff members who performed the tests, the mechanisms through which staff members received training in spirometry, or whether tests were performed according to accepted American Thoracic Society criteria.
Our findings suggest that important shortcomings in spirometric testing are commonplace among pediatric primary care providers, especially among pediatricians. This study highlights physician deficiencies in interpreting spirometric results, as well as perceived barriers to incorporating spirometric testing in clinical practice. To realize the potential benefits of spirometric testing among children with asthma, promotion of spirometry training programs will need to be a priority. Training programs should be especially important for pediatric physicians and their staff members, not only to promote widespread use of spirometry but also to ensure correct interpretation of results. Widespread implementation of the NAEPP guidelines for spirometry likely would require a major initiative to address common barriers experienced by pediatric primary care providers. Future studies should consider the effectiveness of different modalities of spirometry training, especially for pediatric primary care settings.
This work was supported by the Blue Cross Blue Shield of Michigan Foundation.
- Accepted June 9, 2010.
- Address correspondence to Kevin J. Dombkowski, DrPH, MS, University of Michigan, Department of Pediatrics and Communicable Diseases, Division of General Pediatrics, 300 N. Ingalls St, Ann Arbor, MI 48109-5456. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- CI =
- confidence interval •
- NAEPP =
- National Asthma Education and Prevention Program •
- OR =
- odds ratio
- 1.↵National Asthma Education and Prevention Program. Expert Panel Report 3: Guidelines for the Diagnosis and Management of Asthma. Bethesda, MD: National Institutes of Health; 2007. Publication 08–4051
- Miller MR,
- Hankinson J,
- Brusasco V,
- et al
- Eid N,
- Yandell B,
- Howell L,
- Eddy M,
- Sheikh S
- Fuhlbrigge AL,
- Weiss ST,
- Kuntz KM,
- Paltiel AD
- 25.↵National Institute for Occupational Safety and Health. Spirometry Training Program. Spirometry in the occupational setting. Available at: www.cdc.gov/niosh/topics/spirometry/training.html. Accessed July 26, 2010
- Copyright © 2010 by the American Academy of Pediatrics