s55

TELEMEDICAL EDUCATION:

TEACHING SPIROMETRY ON THE INTERNET

Esther H. Lum and Thomas J. Gross

Division of Pulmonary, Critical Care, and Occupational Medicine, Department of Internal Medicine,University of Iowa Hospitals and Clinics, Iowa City, Iowa 52242

Advances in portable equipment have led to routine spirometry testing outside of

formal pulmonary function laboratories. Practitioners ordering these tests are not

formally trained in spirometry interpretation. Providing effective off-site training

can be challenging. Our objective was to develop a remotely accessible computer-based

tutorial for teaching spirometry interpretation to nonpulmonologists. We designed an

educational module that was accessible via the Internet and tested by 65 medical trainees

at a major university medical center. In addition, the module was posted within the

Virtual Hospital on the World Wide Web. Increases in spirometry interpretative skills

were assessed using pre- and post-tests submitted electronically. The spirometry module

significantly improved spirometry interpretation by nonspecialist trainees. This improve-

ment included a broad increase in knowledge base and was observed independent of

training level and prior spirometry reading experience. We conclude that computer-

based tutorials can effectively train off-site practitioners in spirometry interpretation.

This technology allows for the dissemination of educational material from a central site of

expertise and provides a valuable adjunct to limited teaching resources.

AM. J. PHYSIOL. 276 (ADV. PHYSIOL. EDUC. 21): S55S61, 1999.

Key words: pulmonary physiology; pulmonary function tests; Virtual Hospital; medical

education

Pulmonary function tests (PFT) are among the firstdiagnostic studies employed in the evaluation of sus-pected lung disease. Indeed, the measurement of expiredairflow by simple spirometry is invaluable in manyclinical settings including the diagnosis and manage-ment of bronchial asthma, evaluation of chronic cough,and quantification of disability impairment. Recentinnovations in PFT equipment have made portablespirometry devices reliable and accurate (19). This hasallowed for the increasing use of simple spirometry in theprimary care setting. Thus spirograms can now beperformed routinely outside of traditional PFT labora-tories in outpatient office practices, industrial clinics,and community hospital emergency rooms (13, 19).

The practitioners ordering these tests, however, maynot have had formal training in the performance orinterpretation of spirometry. Whereas automated inter-pretation programs are available, these rigid algo-rithms cannot, at present, fully replace an experi-enced reader (3). Educating primary caregivers inspirometry interpretation could lead to more appropri-ate test acquisition, data utilization, and specialtyreferral (4, 15, 16). Providing convenient, effectiveeducational opportunities for these off-site practitio-ners can be a daunting task. Increased personal andinstitutional access to the information superhighway,the Internet, offers new opportunities for the remotedelivery of computer-aided instruction (7).

I N N O V A T I O N S A N D I D E A S

1043 - 4046 / 99 $5.00 COPYRIGHT r 1999 THE AMERICAN PHYSIOLOGICAL SOCIETY

VOLUME 21 : NUMBER 1 ADVANCES IN PHYSIOLOGY EDUCATION JUNE 1999

S55

The University of Iowa has developed and institutedan electronic medical education forum, The VirtualHospital (11). This multimedia, integrated teachingpackage is accessible through the Internet (http://indy.radiology.uiowa.edu/) and is also available toUniversity of Iowa-affiliated clinical outreach centersvia a statewide telemedicine network (17). We havedeveloped a computer-based educational module totest whether such a teaching network could be usedto effectively improve spirometry interpretation skillsfor off-site nonpulmonologists.

METHODS

Educational module. Using published American Tho-racic Society (ATS) guidelines, we designed a com-puter-based tutorial on the interpretation of simplespirometry (1). In addition to stressing recognition ofspecific disease patterns, standards for assessing spiro-gram quality were also addressed (2). To ensure thatour recommendations were consistent with standardpractice, the guidelines were reviewed and approvedby the University of Iowa Hospitals and Clinics PFTlaboratory director as well as an outside expertreviewer (Dr. William Eschenbacher, Baylor Collegeof Medicine, Houston). Textual content was com-posed on the authors personal computer and thentransferred by electronic mail to the Virtual Hospitallibrarian for conversion into hypertext markup lan-guage (HTML) files. Representative spirometry trac-ings were obtained from clinical spirograms per-formed at the University of Iowa PFT laboratory.These figures were scanned into Adobe Photoshopv3.0 (Adobe Systems, San Jose, CA) using a Hewlett-Packard ScanJet IIc and inserted within the textdocument. To aid reader visualization, a normal volun-teer (E. H. Lum) performing spirometry and theflow-volume loop accompanying this effort were videorecorded (Sony Handycam, 8 mm). These video im-ages were digitized and inserted with hot text buttonswithin the body of the spirometry module.

Spirometry interpretation testing. A series of 12spirograms representing the entire range of materialreviewed in the educational module was collected.For each, a brief clinical scenario was composeddetailing a fictitious patients age and sex and thereason for ordering the patients spirogram. A series ofnine possible interpretations accompanied the spiro-

grams. The spirograms and clinical identifiers wereused as stems for R-type multiple matching questionsusing the same list of interpretations as possibleanswers for each question (5). A lead-in instructed thestudents to choose the single best interpretation andstated that each answer could be used once, morethan once, or not at all. The spirograms and thedesignated correct answers were reviewed by clinicalexperts in PFT interpretation (see above). Thesequestions were grouped as a pre-test that was takenbefore reading the educational module. The sameseries of spirograms was relabeled with differentclinical vignettes, randomly reordered, and presentedwith the same set of potential answers as a post-testimmediately on completion of the educational mod-ule. Completion of each test triggered submission ofthe results to a dedicated electronic mailbox. Feed-back on performance was provided via electronic mailto those who requested it.

Subjects were recruited from among the medicalstudents, interns, and more senior house staff rotatingthrough the Department of Internal Medicine accord-ing to a protocol approved by the University of IowaInstitutional Review Board. Subjects were divertedfrom scheduled noon conferences to an educationalresource center complete with multiple personalcomputer stations that provided Internet access withbookmarks directing them to the Virtual Hospital.After a brief introduction, participants were free tocomplete the pre-test, module, and post-test at theirown pace with the option of submitting their testscores under pseudonyms. Those who completed theexercise were compensated with a free meal.

Remote distribution of spirometry module. A sitewas created on the Virtual Hospital that contained thepre-test, module, and post-test. This site was postedwithin the Pulmonary Core Curriculum section with-out specific advertisement or promotion. Respon-dents to the module who submitted the pre-testand/or post-test were asked to identify themselves bylevel of training and/or health care occupation using adrag-down menu. Those who provided electronicmailing addresses were given their test results andexplanations of any incorrect answers.

Data analysis. Test scores were reported as thenumber correct out of a possible score of 12. The time



S56

required to read the module and submit post-testresults was derived from log-on records and recordedin whole minutes. Pre- and post-test scores wereanalyzed using paired, two-tailed t-tests (MicrosoftExcel 5.0 for the Macintosh) with a P value of ,0.05considered a significant difference between popula-tion means. To limit the influence of outlying values,data were also examined using median scores.

RESULTS

The spirometry module improved spirogram inter-pretation by trainees. The spirometry module wasadministered to 65 participants, including 20 third-year medical students, 17 interns, and 28 senior housestaff. The test score results followed a normal distribu-tion (Fig. 1) with scores improving from 6 6 0.3(mean 6 SE) on the pre-test to 8 6 0.3 correct answersout of 12 on the post-test (P , 0.001). Furthermore, themedian test score increased from 5 to 8 correct answersout of a possible 12, demonstrating that the increasein post-test mean score reflected overall group im-provement and not simply individual high scores.

The spirometry module improves test perfor-mance at all levels of training. We are interested indeveloping educational tools for off-site teaching that

are useful for both those in training as well as primarycaregivers in practice. Therefore, the change in pre-and post-test scores was examined for each traineesubgroup to determine how previous knowledge baseinfluenced the effectiveness of the module. At alllevels of training, post-test scores increased signifi-cantly (Fig. 2). The greatest gain in scores wasobserved among the medical students, who alsomanifested the lowest average pre-test scores as ex-pected on the basis of their limited experience withthe clinical use of spirometry (Fig. 2). Subgroups withhigher pre-test scores and, presumably, more experi-ence with interpreting spirometry, still showed signifi-cant improvement in post-test scores (Fig. 2). Simi-larly, nearly all individuals improved their post-testscores whether the pre-test score was above or belowthe median for their respective trainee subgroup (Fig.3). Thus spirometry interpretive performance improvedirrespective of training level or prior experience.

The spirometry interpretative skills improvedacross a wide range of topics. The observed im-provement in spirometry interpretation skills couldhave resulted from learning a few novel facts or from amore global educational enrichment. Analysis of indi-vidual test answers showed that for nearly everyquestion the percentage of correct responses in-creased (Fig. 4). The greatest gains appeared to

FIG. 1.The spirometry module improves spirogram interpre-tative skills. The spirometry module, with pre- andpost-test assessments, was administered to 65 partici-pants. Test scores are distributed normally with im-provement from 6 6 0.3 correct responses on thepre-test (mean 6 SE) to 8 6 0.3 correct responses outof 12 total on the post-test (P F 0.001). Median scorealso increased from 5 to 8 correct responses, confirm-ing that improvement occurred across much of thegroup.

FIG. 2.Spirometry interpretation skills improve at all levelsof training. Test scores were subgrouped by partici-pant training level. Data are displayed as median no.of correct responses out of a possible 12; n 5 no. ofparticipants. Note that the greatest improvement inpost-test scores was observed in the group with thelowest pre-test values (medical students); however,significant improvement was observed for all 3 sub-groups (*P F .001, post-test vs. pre-test median).



S57

involve learning the grading schema for airways ob-struction, recognizing variable upper airway obstruc-tion, and identifying uninterpretable tests. Overall,participation in the module improved spirometryinterpretation across the entire range of disorderstested.

The spirometry module was brief and simple touse. To model the experience of a remote user, ourtest group gained access to the spirometry moduleusing the University of Iowas Internet provider. Thetime required to read the module and complete thepost-test ranged from 25 to 30 min; those participantsat advanced training levels required less time (medicalstudents 30 6 2 min, senior house staff 25 6 2 min).During the test sessions, few problems were encoun-tered using or accessing the module, and most partici-pants did not require additional directions. Thus thisself-directed educational tool was effective, time effi-cient, and user friendly.

The spirometry module was effectively dissemi-nated via the Internet. The spirometry module wasincorporated within the Pulmonary Core Curriculum

FIG. 3.The spirometry module improved interpretive perfor-mance independent of prior experience. Pre- andpost-test scores are shown for intern subgroup as no.of correct responses out of a possible 12, where AQindicate individual participants; arrowheads depictpre- and post-test median scores for entire interngroup. All but 2 participants (A and C) had an im-proved post-test score after completing the spirom-etry module. Individual post-test scores improvedwhether the associated pre-test score was initiallyabove or below the group median. Similar findingswere observed for other groups tested (data notshown).

FIG. 4.Computer-based tutorial produced broad improvement in spirometryreading skills. Analysis of submitted answers for individual test questionsshowed that the percentage of correct responses increased for nearlyevery question tested. The largest increase in correct answers appeared toinvolve the learning of classification schema for grading airways obstruc-tion (questions 1, 4, 10, and 12), recognizing variable upper airwayobstruction (UAO; question 11), and identifying uninterpretable studies(questions 2 and 9). mod, Moderate; r/o, rule out.



S58

of the Virtual Hospital. The module was posted for 12wk without specific advertisement. Despite this lackof solicitation, the site received ,4,000 hits repre-senting the number of times the module was ac-cessed. A pre-test was submitted by 122 participantswho identified themselves as medical students (53%),house staff (12%), respiratory therapists (11%), andothers such as community physicians, nurses, andnon-health care workers. From those who offeredinformation up front and/or responded to follow-upelectronic mail, we were able to identify participantsfrom Canada, the United Kingdom, Italy, Malaysia,Brazil, the Hawaiian Islands, and numerous siteswithin the continental United States. We received 17completed pairs of pre- and post-tests. All 17 respon-dents increased their scores, with a median improve-ment of 2 correct responses (range 1 to 4). Thus, withthe use of an existing electronic teaching network, thespirometry module was internationally displayed andaccessed. For those motivated to submit pre- andpost-tests, use of the module improved spirometryinterpretation.

DISCUSSION

As spirometry becomes more readily available outsideof traditional pulmonary function laboratories, it willmore frequently be ordered by caregivers who maynever have been formally trained in proper interpreta-tion. This lack of instruction combined with anemphasis on shifting diagnostic evaluations away fromspecialized centers could lead to both misdiagnoses aswell as missed diagnoses. Indeed, the need for im-proved understanding of PFT indications and interpre-tation has been documented in other studies. A reviewof PFT ordering patterns by primary care internistsand house staff at a community hospital found that upto two-thirds of the studies requested were not fullyappropriate (16). Another study estimated that nonpul-monologists misinterpret spirometry up to one-thirdof the time with errors in disease classification,overreading of normal spirograms, and misreading ofabnormal studies that illustrated restrictive defects orupper airway abnormalities (15). These authors sug-gest that standards for interpreting spirometry need tobe more readily available to primary care physicians(15). Similarly, a recent Canadian study of PFT labora-tories in British Columbia found marked interobservervariability in interpretation of standard spirograms.This variability was attributed primarily to failure to

adhere to ATS guidelines (4). Thus there is an educa-tional deficit among primary care providers in recog-nizing proper indications for and the correct interpre-tation of PFT that could negatively impact the deliveryof efficient, cost-effective health care.

In response to this educational need, we have de-signed a computer-based tutorial to teach spirometryinterpretation that is brief, effective, and widely acces-sible via the Internet. The criteria and guidelines setforth by the ATS are emphasized in a user-friendly,multimedia format. Our analysis demonstrates that thespirometry module is educationally effective. Partici-pants completing the module improved their ability tocorrectly interpret spirograms (Fig. 1). This improvedperformance was observed among all participantsindependent of level of training or prior experience inreading PFT (Figs. 2 and 3). Interpretative skillsimproved across the range of patterns covered by themodule, demonstrating a broad-based increase inknowledge beyond the acquisition of a few novel facts(Fig. 4). Furthermore, the spirometry module can becompleted with minimal time commitment (,30 min)and little supplementary instruction, which will hope-fully encourage additional sessions as needed.

The durability of the knowledge gained throughparticipation in the spirometry module was not as-sessed. The effectiveness of this computer-based tuto-rial was not directly compared with that of moretraditional printed texts or standard lectures on spirom-etry interpretation. Numerous other studies, however,have confirmed the effectiveness of similar educa-tional tools for teaching radiograph interpretation,pulmonary auscultation, pediatric pulmonary dis-eases, and well-newborn care protocols (8, 1013).Many of the participants in these studies preferredcomputer-based techniques over more traditional de-vices with similar educational efficacy. We envisioncomputer-assisted learning serving as a welcome andeffective adjunct to more traditional curricula.

Certainly, computer-assisted learning has been inte-grated into the medical educational arena, and itspotential applications are seemingly limitless (7, 12,14, 23). However, the real challenge lies in disseminat-ing these educational opportunities to students andpractitioners who may not have immediate physical



S59

access to the training center. Since computers enteredthe medical workplace, their use has become increas-ingly routine, in both the office and the hospital. Afterthe spirometry module was posted on-line within theVirtual Hospital, off-site trainees and primary carepractitioners from various countries around the worldaccessed the tutorial. Pre- and post-tests submitted viaelectronic mail were received, processed, and re-sponded to without further user costs above those forInternet provider access. Thus an effective educa-tional intervention was distributed internationally us-ing technology that allows for central quality controland updating, around-the-clock access, and, if auto-mated, immediate feedback to the participant.

Our study highlights several areas of concern thatmust be addressed for remotely accessible computer-based learning to realize its full potential. First, thequality of presented material must be assured. Al-though we incorporated the central features of theATS recommendations for spirometry standardizationinto the module, clinical interpretative skills are sub-ject to personal biases of the teacher. To maintain thequality of its material, the Virtual Hospital has initiateda peer review system akin to that used by standardpublished journals. All material is confidentially cri-tiqued on-line and submitted by electronic mail tominimize response time; reviews are repeated withscheduled regularity to ensure that material remainscurrent (Dr. Michael Peterson, Virtual Hospital Edito-rial Board, personal communication). Second, wereceived several comments noting great variability inthe quality of displayed graphics and response timesfor viewing video images. Thus the usefulness of amultimedia educational tool may be limited by theequipment available to the end user. This limitationstresses the importance of striving to find applicationsthat function across multiple-user platforms. Finally,although we were pleased to receive nearly 4,000visits at the spirometry module site, completed testpairs represented ,1% of the total audience. To fullyengage the target audience, additional carrot-and-stickincentives such as continuing medical education cred-its or special certification may be required (6, 9, 22).Toward this end, the spirometry module has recentlybeen incorporated into a medical student mini-course along with other similar modules on acid-base interpretation and respiratory pathophysiology.In addition, the module is also available as a continu-

ing medical education offering through the VirtualHospital (http://www.vh.org/Providers/Simulations/Spirometry/SpirometryHome.html).

In summary, we used a computer-based educationalintervention to effectively teach spirometry interpreta-tion to nonpulmonologists. With the use of the VirtualHospital as a central repository, this information wasdistributed to trainees and primary caregivers all overthe world. This teaching service is now provided tomany off-campus referral centers and outreach clinicsas part of a University of Iowa information network.Soon such tutorials may be linked to computerizedpulmonary function reports, facilitating combineddistribution of clinical data and relevant educationalmaterials. Our preliminary study is not intended toprove that electronic educational tools are necessaryor superior to current teaching practices. However, astime and resources for medical education continue toshrink, computer-aided instruction via the Internetoffers promising, cost-effective adjuncts to traditionalmodes of education. In the future, we anticipategreater development and implementation of similareducational interventions.

We thank Teresa Knutson-Choi, Nola Riley, and the Virtual Hospitalsupport staff for expert technical assistance.

This work was supported, in part, by the American Lung Associa-tion of Iowa and a grant from the National Library of Medicine.

Address for reprint requests and other correspondence: T. J. Gross,Room C33-GH, Div. of Pulmonary, Critical Care, and OccupationalMedicine, Univ. of Iowa Hospitals and Clinics, Iowa City, IA 52242(E-mail: [email protected]).

Received 23 July 1998; accepted in final form 10 February 1999.

References

1. American Thoracic Society. Lung function testing: selectionof reference values and interpretative strategies. Am. Rev.Respir. Dis. 144: 12021218, 1991.

2. American Thoracic Society. Standardization of spirometry,1994 update. Am. J. Respir. Crit. Care Med. 152: 11071136,1995.

3. Banks, D. E., M. L. Wang, L. McCabe, M. Billie, and J.Hankinson. Improvement in lung function measurementsusing a flow spirometer that emphasizes computer assessmentof test quality. J. Occup. Environ. Med. 38: 279283, 1996.

4. Broder, S. L., G. Copland, J. Berkowitz, and J. Road.Inter-observer variability of pulmonary function test interpreta-tion (Abstract). Am. J. Respir. Crit. Care Med. 155: A71, 1997.

5. Case, S. M., and D. B. Swanson. Constructing written testquestions for the basic and clinical sciences. [On-line] NationalBoard of Medical Examiners. http://www.nbme.org/new.ver-sion/item.htm [1999, Jan. 28]



S60

6. Chao, J. Continuing medical education software: a compara-tive review. J. Fam. Pract. 34: 598604, 1992.

7. Chodorow, S. Educators must take the electronic revolutionseriously. Acad. Med. 71: 221226, 1996.

8. DAlessandro, M. P., J. R. Galvin, W. E. Erkonen, M. A.Albanese, V. E. Michaelsen, J. S. Huntley, R. M. McBurney,and G. Easley. The instructional effectiveness of a radiologymultimedia textbook (Hyperlung) versus a standard lecture.Invest. Radiol. 28: 643648, 1993.

9. Davis, D. A., M. A. Thompson, A. D. Oxman, and B.Haynes. Evidence for the effectiveness of CME. JAMA 268:11111117, 1992.

10. Desch, L. W., M. T. Esquivel, and S. K. Anderson. Compari-son of a computer tutorial with other methods for teachingwell-newborn care. Am. J. Dis. Child. 145: 12551258, 1991.

11. Galvin, J. R., M. P. DAlessandro, W. E. Erkonen, T. A.Knutson, and D. L. Lacey. The Virtual Hospital: a newparadigm for lifelong learning in radiology. Radiographics 14:875879, 1994.

12. Glowniak, J. V. Medical resources on the Internet. Ann.Intern. Med. 123: 123131, 1995.

13. Guntupalli, K. K., V. Bandi, C. Sirgi, C. Pope, A. Rios, andW. Eschenbacher. Usefulness of flow-volume loops in emer-gency center and ICU settings. Chest 111: 481488, 1997.

14. Headrick, L., A. Kaufman, P. Stillman, L. Wilkerson, andR. Wighton. Teaching and learning methods for new generalistphysicians. J. Gen. Intern. Med. 9: S42S49, 1994.

15. Hnatiuk, O., L. Moores, T. Loughney, and K. Torrington.Evaluation of internists spirometric interpretations. J. Gen.Intern. Med. 11: 204208, 1996.

16. Khatri, K., R. Kaufman, and W. Baigelman. Utilization ofpulmonary function tests by primary care internists in acommunity hospital. Am. J. Med. Qual. 9: 4953, 1994.

17. Kienzle, M. G., D. Curry, J. Galvin, E. Hoffman, E. Holtum,L. Shope, J. Torner, and D. Wakefield. Iowas nationallaboratory for the study of rural telemedicine: a description of awork in progress. Bull. Med. Libr. Assoc. 83: 3741, 1995.

18. Mangione, S., L. Z. Nieman, and E. J. Gracely. Comparisonof computer-based learning and seminar teaching of pulmonaryauscultation to first-year medical students. Acad. Med. 67:S63S65, 1992.

19. Rebuck, D. A., N. A. Hanania, A. D. DUrzo, and K. R.Chapman. The accuracy of a handheld portable spirometer.Chest 109: 152157, 1996.

20. Santer, D. M., V. E. Michaelsen, W. E., Erkonen, R. J.Winter, J. C. Woodhead, J. S. Gilmer, and M. P.DAlessandro. A comparison of educational interventions:multimedia textbook, standard lecture, and printed textbook.Arch. Pediatr. Adolesc. Med. 149: 297302, 1995.

21. Sestini, P., E. Rensoni, M. Rossi, V. Beltrami, and M. Vagli-asindi. Mulitmedia presentation of lung sounds as a learning aid formedical students. Eur. Respir. J. 8: 783788, 1995.

22. Thompson, W. L., J. P. Dyke, and E. Buonocore. Using theWorld-Wide Web to train and certify physicians in the safe useof fluoroscopy. AJR Am. J. Roentgenol. 166: 12631264, 1996.

23. Weiler, R. M. Creating a virtual materials and resources indexfor health education using the world wide web. J. Sch. Health66: 205209, 1996.



S61

s55

Documents