0
Special Feature |

A Systematic Review of the Diagnosis of Occupational Asthma* FREE TO VIEW

Jeremy Beach, MBBS, MD; Kelly Russell, BSc; Sandra Blitz, MSc; Nicola Hooton, BSc; Carol Spooner, MSc; Catherine Lemiere, MD, MSc; Susan M. Tarlo, MB, BS, FCCP; Brian H. Rowe, MD, MSc, FCCP
Author and Funding Information

*From the Departments of Public Health Sciences (Dr. Beach) and Emergency Medicine (Dr. Rowe, Ms. Blitz, and Ms. Spooner), and Capital Health Evidence Based Practice Centre (Ms. Russell and Ms. Hooton), University of Alberta, Edmonton; Department of Chest Medicine (Dr. Lemiere), University of Montreal, Montreal; and Department of Medicine and Public Health Sciences (Dr. Tarlo), University of Toronto, Toronto, Canada.

Correspondence to: Brian Rowe, MD, MSc, Department of Emergency Medicine, 1G1.43 WMC, University of Alberta, 8440-112 St, Edmonton, AB, Canada, T6G 2B7; e-mail: brian.rowe@ualberta.ca



Chest. 2007;131(2):569-578. doi:10.1378/chest.06-0492
Text Size: A A A
Published online

Background: This study systematically reviews literature regarding the diagnosis of occupational asthma (OA) and compares specific inhalation challenge (SIC) testing with alternative tests.

Methods: Electronic databases and trials registries were searched; additional references were identified from bibliographic searches of included studies, hand searches of conferences, and author contacts. Various study designs (clinical trials, cohorts, cross-sectional, or case series) were included involving workers with suspected OA. All diagnostic tests were compared to a “reference standard,” and two researchers independently extracted 2 × 2 data. Pooled sensitivities and specificities (95% confidence intervals [CIs]) were derived.

Results: Seventy-seven studies were included. For high molecular weight (HMW) agents, the nonspecific bronchial provocation (NSBP) test, skin-prick test (SPT), and serum-specific IgE had sensitivities > 73% when compared to SIC. Specificity was highest for specific IgE vs SIC (79.0%; 95% CI, 50.5 to 93.3%). The highest sensitivity among low molecular weight asthmagens occurred between combined NSBP and SPT vs SIC (100%; 95% CI, 74.1 to 100%). When compared to SIC, specific IgE and SPT had similar specificities (88.9%; 95% CI, 84.7 to 92.1%; and 86.2%; 95% CI, 77.4 to 91.9%, respectively). For HMW agents, high specificity was demonstrated for positive NSBP tests and SPTs alone (82.5%; 95% CI, 54.0 to 95.0%) or when combined with specific IgE (74.3%; 95% CI, 45.0 to 91.0%) vs SIC. Sensitivity was somewhat lower (60.6% and 65.2%, respectively).

Conclusions: In appropriate clinical situations when SIC is not available, the combination of a NSBP test with a specific SPT or specific IgE may be an appropriate alternative to SIC in diagnosing OA. While positive results of single NSBP test, specific SPT, or serum-specific IgE testing would increase the likelihood of OA, a negative result could not exclude OA.

Figures in this Article

Occupational asthma (OA) is characterized by airway inflammation, bronchoconstriction, and airway hyperresponsiveness in response to workplace exposures. More than 250 asthmagens have been implicated and identified as causative agents in the development of OA, and new causes are identified each year.1In developed countries, OA is among the most prevalent occupational lung diseases.2A review3 of 43 attributable-risk estimates from 19 counties found the attributable risk of OA to be approximately 9% (interquartile range, 5 to 19%).

The majority of reported OA is OA with latency. Latency periods are observed in all known instances of immunologically mediated asthma, even though the immunologic mechanism may not yet have been clearly identified.2 When a worker has OA with latency, results of specific inhalation testing with the causal agent will usually be positive, and often an immunologic response is achieved with specific skin-prick tests (SPTs) and/or IgE testing, particularly for high molecular weight (HMW) asthmagens. The immunologic responses associated with OA from many low molecular weight (LMW) agents have yet to be fully characterized.2 OA can occur without a prior latent period of exposure. Probably the best known example of this is reactive airways dysfunction syndrome.

While a specific inhalation challenge (SIC) test is considered the reference standard test for OA, a definitive diagnostic test does not exist. False-negative results for SIC can occur when a worker with OA is exposed to more than one asthmagen in the workplace and is challenged with the incorrect asthmagen.4Also, SIC is not widely available, as facilities are confined to a few academic centers in large metropolitan areas.5

There are several other techniques, used in isolation or in combination, that can be used to diagnose OA. In addition to the clinical history or questionnaire, diagnostic tests other than SIC include single and serial nonspecific bronchial provocation (NSBP) challenge, serial lung function testing, specific immunologic testing, exhaled nitrous oxide, and sputum induction, among others. Algorithms have been proposed to aid in establishing a diagnosis of OA.2

Our objective was to systematically gather the existing evidence on diagnostic tests used for OA to determine which techniques are effective in determining a diagnosis. Specifically, we sought to determine the best diagnostic approach for a patient with suspected OA, and to identify when SIC testing could provide additional useful diagnostic information.

A detailed description of the methods can be found at http://www.ahrq.gov/clinic/epcix.htm. A condensed version is presented below.

Literature Search

Appropriate search terms were adapted to search the following electronic resources: MEDLINE, EMBASE, Dissertation Abstracts, Expanded Academic, National Agricultural and Safety Database, CINHAL, Biological Abstracts, Agricola, and trial registries (http://clinicaltrials.gov/; http://www.centerwatch.com/; http://www.cochrane.org/index0.htm; http://www.controlledtrials.com/; http://www.update-software.com/National/; http://www.trialscentral.org/). Web of Science was searched by tracking the most sentinel articles forward. The search was not limited by language or publication status and is considered current until February 2004.

Prominent authors were contacted regarding any missing studies, and the reference lists of all included articles were reviewed. Relevant conference proceedings were hand searched for the years 2001 to 2003.

Study Selection

Each title and, when available, abstract were independently screened by two reviewers. The references identified as “potentially relevant” and “unclear” were then screened by an occupational medicine specialist with an interest in OA and an asthma researcher; the full text of those articles were retrieved.

Using predetermined inclusion criteria, two reviewers independently assessed the full text of the articles. A study was included if it met the following inclusion criteria: (1) de novo asthma or a previous diagnosis of asthma that was exacerbated by a specific agent in the workplace (not including work-aggravated asthma); (2) had a reference standard (SIC or supervised workplace challenge, immunologic testing, and serial lung function tests or serial measurement of nonspecific airway reactivity, or clinical expert diagnosis and exposure to an asthmagen) and a comparison test (the above tests and/or sputum, metabonomics, nitrous oxide); and (3) results were presented including absolute numbers to construct a 2 × 2 table, sensitivity, specificity, likelihood ratios, cost, time to complete diagnosis, or adverse effects. Disagreement among reviewers was resolved by discussion and consulting a third party as needed.

We included research from appropriate clinical settings and excluded studies that only assessed screening processes for workplace respiratory symptoms. In cases in which multiple publications involving the same or a portion of the same workers were identified, the most recent article or the largest cohort was selected and any additional, unique information from previous publications was incorporated.

Quality Assessment

Assessment of the methodologic quality was completed independently by two reviewers. The methodologic quality of a diagnostic study was assessed using a quality tool based on empirical research of Lijmer et al6 examining biases in diagnostic studies. Empirically validated questions included study design, blinding of measurement(s), use of an appropriate reference standard, an adequate description of the reference standard, test(s), population, and the occurrence of differential reference bias. In addition, information regarding the occurrence of partial verification bias, timing of data collection, method of worker selection, reporting of interrater reliability, and method of reporting results was also captured. Also, the funding source and whether medication was ceased (or attempted) prior to testing were recorded for each study.

Data Extraction

Data were extracted by one reviewer using standardized forms and were checked for completeness and accuracy by a second reviewer. Extracted information included data regarding the study population characteristics, diagnostic tests, and results (eg, sensitivity, specificity, likelihood ratios, etc). For each diagnostic test, the methodology, timing, inclusion/exclusion criteria, and definition of a positive test result were extracted. Additional outcomes included cost of diagnosis, time to complete diagnosis, and presence of adverse events.

Data Analysis

Standard 2 × 2 tables for each comparison test or combination of tests were generated. Sensitivity and specificity were calculated for each study using standard formulas. When sensitivity and specificity could be calculated for more than one cut-off value or definition for the same diagnostic test, we included the table that produced the highest test efficiency defined as the proportion of correctly identified patients (reference standard positive and reference standard negative) by the comparison test. Results were pooled using the inverse variance method for random effects to calculate an overall estimate of sensitivity and specificity (Meta-Test 0.9, Meta package for R; J. Lau; New England Medical Center; Boston, MA). All results are presented with 95% confidence intervals (CIs). Results were only pooled for studies of similar molecular weight (HMW, LMW, or mixed) asthmagens. Sensitivity/specificity pairs were plotted in receiver operating characteristic space.7 Positive and negative likelihood ratios were also estimated. Heterogeneity was assessed visually.

Search Strategy

The search strategy and complete reference list of included studies are available on the Agency for Healthcare Research and Quality (AHRQ) Web site cited above. The literature search resulted in 7,112 studies, and 487 unique citations were identified as potentially relevant. Seventy-seven studies were ultimately included; exclusions are shown in Figure 1 .

Description of Included Studies

The 77 unique cohorts that were included in the diagnosis review are described in Table 1 . SIC was the most commonly cited reference standard; it was performed in 68 of the studies and most frequently with di-isocyanates. The alternative reference standard identified was “clinical diagnosis.” Over half of the studies (n = 38) included workers exposed to LMW agents. The comparison diagnostic test result most commonly reported was a single NSBP test. Worker characteristics were measured across studies and are described in Table 2 .

Methodologic Quality of Included Studies

The methodologic quality of the included studies is described in Table 3 . Workers were relatively infrequently selected consecutively or randomly, and 21 studies failed to report the selection method; the remaining studies used alternative methods, such as choosing some workers in a factory or some clinic patients. All but 14 studies collected data prospectively; data collection was unclear in four studies.

Differential reference bias was assessed as likely to have occurred in 16 studies, could not be assessed in 9 studies, and thought unlikely in the remainder. Partial verification bias was assessed as likely in 19 studies and could not be excluded in 11 studies. Partial verification bias did not occur in the remaining 47 studies.

Single NSBP Test vs SIC

Thirty-nine studies compared the sensitivity and specificity of single NSBP test to SIC. Among the 10 studies examining HMW workplace exposures, the pooled estimate for sensitivity was 79.3% (95% CI, 67.7 to 87.6%) and for specificity was 51.3% (95% CI, 35.2 to 67.2%). The sensitivity/specificity pairs are plotted in Figure 2 , top left, A. Twenty-four studies reported sufficient data that we could estimate both sensitivity and specificity among patients exposed to LMW agents. The pooled estimate for sensitivity was 66.7% (95% CI, 58.4 to 74.0%) and for specificity was 63.9% (95% CI, 56.1 to 71.0%). The sensitivity/specificity pairs are plotted in Figure 2, top right, B. Five studies reported sensitivity and specificity results for various suspected agents of mixed molecular weights. The pooled estimate of sensitivity was 83.7% (95% CI, 66.8 to 92.9%) and of specificity was 48.4% (95% CI, 25.9 to 71.6%).

Specific SPT vs SIC

Twenty-six studies reported comparisons of specific SPT to SIC. Of the 16 studies that allowed estimation of both sensitivity and specificity for patients exposed to HMW agents, the pooled sensitivity was 80.6% (95% CI, 69.8 to 88.1%), while the pooled specificity was 59.6% (95% CI, 41.7 to 75.3%). The sensitivity/specificity pairs are plotted in Figure 2, center left, C. Five studies reported data allowing estimation of both sensitivity and specificity of SPT of LMW agents; the pooled sensitivity was 72.9% (95% CI, 59.7 to 83.0%), while the pooled specificity was 86.2% (95% CI, 77.4 to 91.9%). The sensitivity/specificity pairs are plotted in Figure 2, center right, D. Among the five studies that included patients exposed to mixed molecular weight agents, the pooled estimates for sensitivity was 63.0% (95% CI, 41.5 to 80.3%) and the pooled estimate of specificity was 59.2% (95% CI, 45.4 to 71.7%).

Serum-Specific IgE vs SIC

Twenty-two studies reported data sufficient to allow comparison of serum-specific IgE with SIC. Sensitivity was higher for the HMW agents studied; the pooled estimate of sensitivity was 73.3% (95% CI, 63.9 to 81.0%), and the pooled estimate of specificity was 79.0% (95% CI, 50.5 to 93.3%). The sensitivity/specificity pairs are plotted in Figure 2, bottom left, E. Among the 11 studies examining LMW agents, the pooled sensitivity was 31.2% (95% CI, 22.9 to 40.8%) and the pooled specificity was 88.9% (95% CI, 84.7 to 92.1%). The sensitivity/specificity pairs are plotted in Figure 2, bottom right, F. Two studies examining a variety of mixed molecular weight agents reported data sufficient to allow estimation of sensitivity and specificity. The pooled estimate of sensitivity was 85.1% (95% CI, 40.3 to 98.0%), and the pooled estimate of specificity was 61.2% (95% CI, 7.0 to 97.1%).

Single NSBP Test and SPT With or Without Serum-Specific IgE vs SIC

When possible, results were combined in union for the most frequently reported comparison tests. That is, results of all tests in combination had to be positive for the combined result to be considered positive; if any result was negative, the combination was considered negative. Four studies examined a single NSBP test and specific SPT in combination among subjects exposed to HMW agents. The pooled estimate of sensitivity was 60.6% (95% CI, 21.0 to 89.9%), while the estimate of specificity was 82.5% (95% CI, 54.0 to 95.0%). Only one study investigated this combination of tests in LMW OA; the sensitivity was 100% (95% CI, 74.1 to 100%) and specificity was 80.0% (95% CI, 49.0 to 94.3%). Three studies of HMW agents yielded results for the combination of a single NSBP test, specific SPT, and serum-specific IgE. The pooled estimate of sensitivity was 65.2% (95% CI, 6.7 to 98.0%) and specificity was 74.3% (95% CI, 45.0 to 91.0%).

Other Tests

Serial NSBP tests were infrequently studied (n = 6) and demonstrated variable results. In the single study examining serial NSBP tests in OA suspected to be due to HMW agents, serial NSBP tests demonstrated 100% sensitivity and specificity in predicting the outcome of SIC.8 The lack of additional evidence precludes firm conclusions. Additional comparisons to SIC are summarized in Table 4 . Information on cost of diagnosis, time to complete diagnosis, and presence of adverse events could not be gathered from the included studies.

Role of Pretest Probabilities

Figure 3 , top left, A, top right, B, and center left, C demonstrate the effect of varying pretest probabilities and the use of the likelihood ratios on the posttest probabilities.

There are a number of diagnostic tests available to clinicians to assist in the diagnosis of OA; however, varying test efficiency, availability, safety profiles, and cost limit their application. For example, of 235 patients with a diagnosis of OA in Ontario, only 43 patients (18%) underwent SICs and 41 patients (17%) underwent SPTs to a work agent; whereas OA in 94 patients (40%) was diagnosed by serial peak expiratory flows (PEFs), 84 patients (36%) underwent an NSBP (methacholine) challenge during a work week, and 132 patients (56%) underwent off-work NSBP challenge.9On many occasions, the reference standard test (ie, SIC) is not available. In addition, SIC is not a perfect test: there is a lack of method consensus, facilities capable of performing SIC are limited, and false-positive and false-negative results can occur.10

This review examined the diagnostic utility of several tests used to diagnose OA, and will hopefully aid clinicians and researchers in selecting diagnostic tests for OA. A positive result from some of the diagnostic tests considered could assist clinicians to rule in OA, although the test may not be considered completely confirmatory when used alone. Further, while a single negative test result would reduce the likelihood of the diagnosis, the reduction would probably not be sufficient to confidently exclude OA.

In the appropriate clinical setting, the data showed that a single NSBP test demonstrated reasonable sensitivity in predicting the outcome of SIC, while specificity was somewhat lower. This may reflect the fact that a positive NSBP test result is usual in symptomatic asthma patients from any cause, and in the context in which SIC is usually performed, ie, in a patient with a history of asthma symptoms worsening at work and improving off work. These findings would support the contention that the first step to diagnosing OA is to diagnose asthma, and this can be supported with a positive NSBP test result. Immunologic testing also appeared useful; SPTs generally showed a higher sensitivity than serum IgE in comparison with SIC, whereas the converse was true for specificity. A clinical “expert” diagnosis (often comprised of many diagnostic tests that may or may not have included SIC) also resulted in high sensitivities for HMW, LMW, and mixed workplace exposure; however, the lower specificity suggested a potential problem with a high proportion of false-positive results. In addition, there were differences in how these tests performed in HMW and LMW agents.

Although the sensitivity and specificity of some of these tests appears to be relatively good, it is important to recognize that these results are produced from highly selected workers from appropriate clinical settings. Generally, they were prescreened, usually by referral to a specialist clinic or a workplace surveillance program documenting exposure to a known agent and a positive response to a screening questionnaire. This process likely increased the pretest probability of OA within the study populations, and may have significant implications when interpreting the estimates of sensitivity and specificity. For example, NSBP, a test often used to help diagnose asthma, had a relatively high sensitivity and specificity in diagnosing OA. These relatively high estimates of sensitivity and specificity may have been generated because the test was applied to workers in whom a screening process had already identified the presence of a likely workplace exposure to an asthmagen. This may be an optimistic estimate of the utility of each test, and may be mainly applicable when applied to a cohort of workers with exposure to a sensitizer. However, in a pulmonary clinic population of asthmatics, the utility may be less because almost all patients would be expected to have positive NSBP results and, thus, the discriminatory value of the NSBP may be lower. Caution should be applied to the results of these estimates when examining workers in the various setting.

When considering these data, it is also important to recognize a number of limitations. When determining the accuracy of any diagnostic test, a reference standard must be available for comparison. The usual reference test for OA is SIC. Although SIC is not always readily available, and the various methodologies and definitions make standardization difficult, there are probably no better alternatives in OA diagnosis at this time. Moreover, while serial PEFs and paired NSBP tests are often used in clinical practice, limited evidence on their diagnostic test characteristics could be identified in the literature.

Not all studies presented data in a form that we could extract and utilize in the metaanalysis. It was not possible to use results that were presented as a difference between mean values of the comparison test when grouped by the reference test result. In other cases, individual patient data were available, but the absence of a referenced cut-off value to define a “positive” test result excluded these results. Consequently, fewer studies than those identified could be included in the pooled analysis.

While SIC was considered the reference standard in the majority of included studies, some studies did not report this result for all patients. Rather, some studies used a clinical “consensus” diagnosis to determine the presence of OA, which may have included SIC; it was usually not clear which patients had undergone SIC and why. This may introduce two forms of bias: differential reference standard bias and partial verification bias.6

As in any systematic review, there is a possibility of publication bias. By missing unpublished unimpressive diagnostic studies, we may be overestimating the diagnostic properties of OA tests. However, a comprehensive search was conducted, authors were contacted, and gray literature was extensively searched. Despite these efforts, we recognize that some studies may exist that we were not able to include, although we do not believe that this would be a large number.

We aimed to consider whether a combination of tests might have sufficiently high sensitivity and specificity to be a suitable alternative to SIC. It would be expected that multiple different positive test results in a patient would result in increased specificity but reduced sensitivity, and that a single positive test result when many different tests were performed would increase sensitivity at the expense of specificity. The highest sensitivity and specificity arose from a combination of a single NSBP test and SPT. Given a high pretest probability of disease, a positive combined test result would virtually confirm OA (posttest likelihood of > 90%). However, negative combined testing results do not appear to provide clinicians with sufficient certainty to rule out OA. It should be noted that the data obtained precluded quantitative evaluation of a number of other sequential testing schemes or combinations of diagnostic tests that have been previously proposed by others2 and were unable to quantify these assumptions.

The figures examining sensitivity/specificity pairs do show substantial heterogeneity between the results of the included studies. For many of the comparison tests, sensitivity and/or specificity ranged from 0 to 100% in the individual studies. Pooled estimates were calculated, but given the heterogeneity within the data we believe they should be interpreted with caution. The disadvantage of pooling sensitivity and specificity separately is that it does not account for the inherent pairing of these two values within a study. We leave it to the clinicians to judge the applicability of the pooled results.

The tests results may be influenced by the time elapsed between assessment and the last exposure to the causative agent or workplace. In the included studies, this information as not always provided nor standardized. We refer the reader to appendix E-1 in the AHRQ report for a complete list of the included studies that documents the substances used in diagnostic testing.

In practice, clinical diagnosis relies heavily on some tests that may not be well characterized but are more readily available than SIC. For example, from one study,9 serial PEFs were used more often than SIC and SPT to diagnose OA, and it has been proposed that serial PEFs be used to confirm the diagnosis of OA.11And in the case of LMW asthmagens for which immunologic testing is not available, these other tests may be the only available alternative. The tests used to diagnose OA also vary based on the region and country. For example, patients in Finland require objective evidence prior to the diagnosis of OA for compensation purposes; thus, SIC is performed in the majority of workers.12 This is not the case for all Canadian provinces9 and the United States.5 For the present, the diagnosis of OA is likely to remain a pragmatic and inexact skill, but these data at least help define the current state of the available evidence with regard to test characteristics. Clearly, additional research could help characterize the validity of the tests further.

Despite limited availability, lack of standardized protocols, and recognized possibilities of false-positive and false-negative results, SIC has been used as the main reference standard for the diagnosis of OA in research publications. In isolation, none of the other diagnostic tests yielded a sufficiently high combination of sensitivity and specificity that they could replace SIC. In the absence of SIC, NSBP testing is probably a pragmatic and readily available alternative test that can be of use within a highly selected population where prescreening by history has occurred (which has presumably identified a likely workplace link). Nonetheless, its sensitivity and specificity alone are insufficiently discriminative to rule in or rule out OA. Adding SPT can enhance the sensitivity and specificity of NSBP testing; many other combination tests have not been evaluated in sufficient detail to provide recommendations.

Priority should be given to ensuring the highest methodologic quality of future research and full reporting of all comparisons conducted to investigate the diagnosis of OA. There is an urgent need for clear comparisons between reference standards (currently SIC) and alternative test approaches, performed independently, and reported using standardized diagnostic test methods, while also assessing the cost of diagnosis, time to complete diagnosis, and presence of adverse events

SIC is not widely available, and the presence of OA is often determined by non-reference standard testing. Although this review has considered the sensitivity and specificity of various tests in comparison to SIC, it has not identified data relating to the resources required for the various types of testing. A prospective cost-benefit analysis of the non-reference standard procedures might also be conducted to determine the costs and benefits associated with using tests other than SIC to determine the presence or absence of OA.

Abbreviations: AHRQ = Agency for Healthcare Research and Quality; CI = confidence interval; HMW = high molecular weight; LMW = low molecular weight; NSBP = nonspecific bronchial provocation; OA = occupational asthma; PEF = peak expiratory flow; SIC = specific inhalation challenge; SPT = skin-prick test

Dr. Beach participated in writing the proposal, provided methodologic and content expertise, and participated in writing and editing the manuscript. Ms. Russell planned, oversaw, and participated in all steps of the systematic review process and in writing and editing the proposal and manuscript. Ms. Blitz performed all statistical analysis, participated in most steps of the systematic review process, and participated in writing and editing the manuscript. Ms. Hooton and Ms. Spooner participated in most steps of the systematic review process and in editing the manuscript. Drs. Lemiere and Tarlo participated in editing the proposal, provided content expertise, and participated in editing the manuscript. Dr. Rowe participated in writing the proposal, provided methodologic and content expertise, and participated in writing and editing the manuscript.

This study was supported by the AHRQ.

The investigating authors acknowledge the following financial support: Dr. Beach is supported by the Faculty of Medicine and Dentistry at the University of Alberta, Edmonton, Alberta; and Dr. Rowe is supported by the Canadian Institutes of Health Research, Canada Research Chairs Program.

The authors have no conflicts of interest to disclose.

Figure Jump LinkFigure 1. Depiction of the study selection process.Grahic Jump Location
Table Graphic Jump Location
Table 1. Description of Included Studies
Table Graphic Jump Location
Table 2. Description of OA Workers
Table Graphic Jump Location
Table 3. Methodologic Quality of Included Studies
Figure Jump LinkFigure 2. All plots show sensitivity/specificity pairs derived from individual studies. Plotting characters are proportional to the sample size. Horizontal lines represent the pooled sensitivity estimate. Vertical lines represent 1-pooled specificity estimates. Top left, A: SIC compared to single NSBP test for HMW compounds. Top right, B: SIC compared to single NSBP test for LMW compounds. Center left, C: SIC compared to SPT for HMW compounds. Center right, D: SIC compared to SPT for LMW compounds. Bottom left, E: SIC compared to specific IgE for HMW compounds. Bottom right, F: SIC compared to specific IgE for LMW compounds.Grahic Jump Location
Table Graphic Jump Location
Table 4. Sensitivity and Specificity of Comparison Tests That Used SIC as a Reference Standard
Figure Jump LinkFigure 3. Nomogram of likelihood ratio among workers with a high pretest probability (75%) [top panels, A], moderate pretest probability (50%) [center panels, B], and low pretest probability (15%) [bottom panels, C] of OA. The line marked “+” indicates the likelihood of disease given a positive test result. The line marked “−” indicates the likelihood of disease given a negative test result.Grahic Jump Location

We thank the American College of Chest Physicians for sponsoring this research through the AHRQ. We are grateful to members of our technical expert panel (Drs. Dilini Vethanayagam, Richard Jones, David Muir) and Capital Health Evidence Based Practice Centre staff (Dr. Terry Klassen, Ms. Lisa Hartling) for providing input on the direction and scope of the review. We appreciate Ms. Ellen Crumley’s expertise in conducting the literature search. We thank Ms. Natasha Wiebe and Ms. Marlene Dorgan for their contribution to the systematic review process.

van Kampen, V, Merget, R, Baur, X (2000) Occupational airway sensitizers: an overview on the respective literature.Am J Ind Med38,164-218. [PubMed] [CrossRef]
 
Chan-Yeung, M, Malo, JL Occupational asthma.N Engl J Med1995;333,107-112. [PubMed]
 
Blanc, PD, Toren, K How much adult asthma can be attributed to occupational factors?Am J Med1999;107,580-587. [PubMed]
 
Chan-Yeung, M, Malo, JL, Tarlo, SM, et al Proceedings of the first Jack Pepys Occupational Asthma Symposium.Am J Respir Crit Care Med2003;167,450-471. [PubMed]
 
Ortega, HG, Weissman, DN, Carter, DL, et al Use of specific inhalation challenge in the evaluation of workers at risk for occupational asthma: a survey of pulmonary, allergy, and occupational medicine residency training programs in the United States and Canada.Chest2002;121,1323-1328. [PubMed]
 
Lijmer, JG, Mol, BW, Heisterkamp, S, et al Empirical evidence of design-related bias in studies of diagnostic tests.JAMA1999;282,1061-1066. [PubMed]
 
Moses, LE, Shapiro, D, Littenberg, B Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations.Stat Med1993;12,1293-1316. [PubMed]
 
Alvarez, MJ, Estrada, JL, Gozalo, F, et al Oilseed rape flour: another allergen causing occupational asthma among farmers.Allergy2001;56,185-188. [PubMed]
 
Tarlo, SM, Liss, G, Corey, P, et al A workers’ compensation claim population for occupational asthma.Chest1995;107,634-641. [PubMed]
 
Tarlo, S Laboratory challenge testing for occupational asthma.J Allergy Clin Immunol2003;111,692-694. [PubMed]
 
Moscato, G, Godnic-Cvar, J, Maestrelli, P Statement on self-monitoring of peak expiratory flows in the investigation of occupational asthma: Subcommittee on Occupational Allergy of European Academy of Allergy and Clinical Immunology.J Allergy Clin Immunol1995;96,295-301. [PubMed]
 
Lemiere, C Evaluation of the patient with occupational asthma: physiologic basis of respiratory disease.2005,727-732 BC Decker. Hamilton, ON, Canada:
 

Figures

Figure Jump LinkFigure 1. Depiction of the study selection process.Grahic Jump Location
Figure Jump LinkFigure 2. All plots show sensitivity/specificity pairs derived from individual studies. Plotting characters are proportional to the sample size. Horizontal lines represent the pooled sensitivity estimate. Vertical lines represent 1-pooled specificity estimates. Top left, A: SIC compared to single NSBP test for HMW compounds. Top right, B: SIC compared to single NSBP test for LMW compounds. Center left, C: SIC compared to SPT for HMW compounds. Center right, D: SIC compared to SPT for LMW compounds. Bottom left, E: SIC compared to specific IgE for HMW compounds. Bottom right, F: SIC compared to specific IgE for LMW compounds.Grahic Jump Location
Figure Jump LinkFigure 3. Nomogram of likelihood ratio among workers with a high pretest probability (75%) [top panels, A], moderate pretest probability (50%) [center panels, B], and low pretest probability (15%) [bottom panels, C] of OA. The line marked “+” indicates the likelihood of disease given a positive test result. The line marked “−” indicates the likelihood of disease given a negative test result.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1. Description of Included Studies
Table Graphic Jump Location
Table 2. Description of OA Workers
Table Graphic Jump Location
Table 3. Methodologic Quality of Included Studies
Table Graphic Jump Location
Table 4. Sensitivity and Specificity of Comparison Tests That Used SIC as a Reference Standard

References

van Kampen, V, Merget, R, Baur, X (2000) Occupational airway sensitizers: an overview on the respective literature.Am J Ind Med38,164-218. [PubMed] [CrossRef]
 
Chan-Yeung, M, Malo, JL Occupational asthma.N Engl J Med1995;333,107-112. [PubMed]
 
Blanc, PD, Toren, K How much adult asthma can be attributed to occupational factors?Am J Med1999;107,580-587. [PubMed]
 
Chan-Yeung, M, Malo, JL, Tarlo, SM, et al Proceedings of the first Jack Pepys Occupational Asthma Symposium.Am J Respir Crit Care Med2003;167,450-471. [PubMed]
 
Ortega, HG, Weissman, DN, Carter, DL, et al Use of specific inhalation challenge in the evaluation of workers at risk for occupational asthma: a survey of pulmonary, allergy, and occupational medicine residency training programs in the United States and Canada.Chest2002;121,1323-1328. [PubMed]
 
Lijmer, JG, Mol, BW, Heisterkamp, S, et al Empirical evidence of design-related bias in studies of diagnostic tests.JAMA1999;282,1061-1066. [PubMed]
 
Moses, LE, Shapiro, D, Littenberg, B Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations.Stat Med1993;12,1293-1316. [PubMed]
 
Alvarez, MJ, Estrada, JL, Gozalo, F, et al Oilseed rape flour: another allergen causing occupational asthma among farmers.Allergy2001;56,185-188. [PubMed]
 
Tarlo, SM, Liss, G, Corey, P, et al A workers’ compensation claim population for occupational asthma.Chest1995;107,634-641. [PubMed]
 
Tarlo, S Laboratory challenge testing for occupational asthma.J Allergy Clin Immunol2003;111,692-694. [PubMed]
 
Moscato, G, Godnic-Cvar, J, Maestrelli, P Statement on self-monitoring of peak expiratory flows in the investigation of occupational asthma: Subcommittee on Occupational Allergy of European Academy of Allergy and Clinical Immunology.J Allergy Clin Immunol1995;96,295-301. [PubMed]
 
Lemiere, C Evaluation of the patient with occupational asthma: physiologic basis of respiratory disease.2005,727-732 BC Decker. Hamilton, ON, Canada:
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

CHEST Journal Articles
CHEST Collections
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543