and Disability among Applicants - NCBI

Occupational

I_mpairment

and Disability among Applicants for Social Security Disability Benefits in Pennsylvania

Eddy A. Bresnitz, MD, MS, Howard Frumkin, MD, DrPH, Lawrence Goldstein, MD, David Neumark PhD, Michael Hodgson, MD, MPH, and Carolyn Needleman, PhD

Introduction l :

F

i:.i :'.:}

..

'.....:.

4- e- ^-

7

:.- :; :;

tyAt

d

i-

It ¢4;,ld

To date, it has been difficult to accurately the extent of workrelated morbidity and mortality in the United States. One reason for this has been the absence of a comprehensive, nationwide, active surveillance system for occupational disease and injury. The absence of such a system also impedes efforts to design, implement, and evaluate interventions to treat and educate affected workers and prevent hazardous exposures in the workplace.1>3 Given this lack, public health officials have instead used a variety of data systems for surveillance purposes,4 including cancer registries5; workers' compensation,67 pension plan, and life insurance records8; death certificates9; records from medical practices'0"'; hospital discharge records'2; court records'3 clinical laboratory reports'4; and records from poison control centers.l5'l6 A patient identified through any one of these systems may serve as an index case, or a sentinel event, for identifying hazardous work conditions and other ill or injured workers.'7 For many workers with occupationally caused illnesses and injuries, the federal Social Security Disability Insurance benefit program serves as an alternative or backup benefit system for workers' compensation.'8 Therefore, this program is another potentially important database estimate

for occupational disease and injury surveillance. Central to an understanding of disability insurance are the concepts of

4fi '-S;

impairment and disability. Impairment, according to the Social Security Administration, is defined as "medically demonstrable anatomical, physiological, or psychological abnormalities."'9 In contrast, disability is defined as

the inability to engage in any substantial activity by reason of any medically determinable physical or mental impairment which can be expected to result in death or has lasted or can be expected to last for a continuous period of not less than 12 months.'9

Disability insurance provides wage replacement and medical benefits to an ill or injured person whose impairment meets the program's stringent disability test. Each state can designate a state agency responsible for reviewing applications and making disability determinations. In Pennsylvania, claimants for disability insurance benefits apply to the Bureau of Disability Determination in the Department of Labor and Industry. In principle, the records of this bureau can serve as a source of information on the impaired and disabled population, potentially including information on workplace Eddy A. Bresnitz is with the Medical College of Pennsylvania, Philadelphia, Pa. Lawrence Goldstein is with the Pennsylvania Bureau of Disability Determination, Harrisburg. At the time of the study, David Neumark and Howard Frumkin were with the University of Pennsylvania, Philadelphia; Dr Neumark is now with Michigan State University, East Lansing, and Dr Frumkin is with Emory University, Atlanta, Ga. Michael Hodgson was with the University of Pittsburgh; he is now with the University of Connecticut, Farmington. Carolyn Needleman is with Bryn Mawr College, Byrn Mawr, Pa. Requests for reprints should be sent to Eddy A. Bresnitz, MD, MS, Division of Occupational and Environmental Health, Department of Community and Preventive Medicine, Medical College of Pennsylvania, 3300 Henry Ave, Philadelphia, PA 19129. This paper was accepted March 8, 1994. Note. The Pennsylvania Department of Health disclaims responsibility for any analyses, interpretations, or conclusions in this article.

November 1994, Vol. 84, No. 11

Work Disability

causes of impairment and disability. However, neither the bureau nor the Social Security Administration store work-related information on applicants in a readily retrievable fashion. Because disability insurance benefits are awarded based on the presence and degree of functional impairment rather than on etiology, the disability application form does not specifically request information on the work relatedness of an individual's medical condition. Instead, the application asks how the applicant's condition affects his or her ability to work; and it inquires into past jobs, including job title, type of business, and dates worked as far back as 15 years from the time of application for benefits. With these data, it is possible to make judgments about whether the applicant's previous employment could have plausibly contributed to current impairment and disability. To assess the extent of workplacerelated disease and injury among disability insurance applicants, then, we analyzed information contained in the bureau's records for 2 months in 1990. This study was conducted as part of a larger report on occupational disease and injury in Pennsylvania.20

Methods Because the work-related information supplied on disability insurance applications, as described above, is not routinely abstracted onto Social Security data tapes, the proportion of past disability applicants in Pennsylvania with workrelated problems could not be readily assessed. As an alternative approach, 240 consecutive first-time applications assigned to a single claims examiner at the bureau (L. Goldstein) were reviewed. The applications were processed in the bureau's Harrisburg branch (one of three) during the 2-month period from February through March 1990. E. Bresnitz and H. Frumkin, two occupational physicians, independently reviewed each case and made determinations ofwork relatedness; there was concurrence on all cases. A case was considered to have a work-related component if there was a clear statement of a workplace illness or injury associated with the impairment (e.g., applicant fell off scaffold at work). Additionally, a case was labeled as work related if the applicant had worked at an occupation with a high likelihood of exposures known or suspected to contribute to the condition of interest, based on existing literaNovember 1994, Vol. 84, No. 1 1

TABLE 1 Work-Related Injuries or Illnesses In 240 Applicants for Social Security Disability Insurance, Pennsylvania, 1990

Case

Diagnosis

1

6

Squamous cell carcinoma of lung Squamous cell carcinoma of pharynx Lumbar disk disease Below-knee amputation Carpal tunnel syndrome and ulnar neuropathies Small cell carcinoma

7 8

Hip fracture Lung cancer

2

3 4

5

9 10

11

Occupation

Yes

Smoker

Welder

Yes

Smoker

Police officer Painter//contractor

Yes Yes

Injured at work Fall from scaffolding at work

Machin(e operator

Yes

Plumbe.r/electrician

Yes Yes Yes Yes Yes

Construction worker

Yes

injury 12

Degenerative joint

Forklift operator

Yes

13 14

Lung cancer T8-9 fracture/dislocation with paraplegia Rotator cuff tendonitis of shoulder Multiple fractures Coal workers' pneumoconiosis

Construction laborer Laborer

Yes Yes

Newspaper deliveryperson Truck driver Coal worker

No

No No

Construction worker Painter

No No

15 16 17 18 19

Lower back pain Cervical disk disease

20

Open comminuted Preassembler at intra-articular fracsteel plant

No

ture

22

Bench press operator Lumbar disk hernia- Office cleaner tion

23

Back pain

Maintenance worker

No

24

Lumbosacral and neck sprain Respiratory symp-

Cook

No

Plastics industry worker

No

Construction

No

Pressman

No

21

25

Eye injury

toms

26 27

Lumbar compression Cervical disk dis-

Comments

Welder

Security guard Worker at naval shipyard Lung cancer Steel welder Below-knee amputa- Painter tion Ill-defined back

Disability Allowance

No No

No information on smoking Injured at work Smoker Smoker Fall from scaffolding caused injuries resuiting in amputation Allowance based on Best's disease of eyes Developed back pain acutely at work Smoker Fall from ladder ...

Injured at work Already receiving black lung benefits Strain on job in 1973 Ceiling painting may have contributed to disk problems While at work, jumped out of second-story window to escape boyfriend Injured at work Lower back pain exacerbated by work activity Symptoms date to a fall at work Fall at work

Respiratory symptoms at work with normal pulmonary function tests Injured at work

ease aSee text for definition of sample selection.

American Journal of Public Health 1787

Bresnitz et al.

TABLE 2-Impairment and Disability, by Organ System, among 240 Pennsylvania Applicants for Social Security Disability Insurance (SSDI), 1990

Cases in Study Sample

Total Applicants with WorkRelated Conditions

SSDI Awards Made for

Disabilities % No.

Organ System

No.

%

Musculoskeletal Special senses Respiratory Cardiovascular Digestive Genitourinary Hematologic Skin Endocrine Multiple organ systems (includes human immunodeficiency virus) Neurologic Neoplastic None established Total

37 6 6 36 2 2 7 1 5 52

15.4 2.5 2.5 15.0 1.0 1.0 2.9 0.0 2.1 21.7

15 2 2 21 1 2 5 0 3 45

25 59 2 240

10.4 24.6 1.0 100.1

19 50 1 166

Applicants with WorkRelated Disabilities Receiving SSDI Awards

No.

%

No.

%

40.5 33.3 33.3 58.3 50.0 100.0 71.4 0.0 60.0 86.5

11 2 2 3 0 0 0 0 0 1

29.7 33.3 33.3 8.3 0.0 0.0 0.0 0.0 0.0 1.9

4 1 1 1 0 0 0 0 0 0

10.8 16.8 16.8 2.8 0.0 0.0 0.0 0.0 0.0 0.0

76 84.7 50.0 69.2

1 6 1 27

4.0 10.2 50.0 11.3

1 6 0 14

4.0 10.2 0.0 5.8

TABLE 3-Distribution and Frequency of Work-Related Conditions and Social Security Disability Insurance (SSDI) Awards Estimated from Study Sample, Pennsylvania, 1990a Projected Numbers (95% Cl)C

Work-Related Projected Applicantsb (95% Cl)C

Organ System

Musculoskeletal 26 020 (22 221, 29 389) 1326 (609, 2867) Special senses 4853 (3226, 7168) Respiratory Cardiovascular 13 046 (10 322,16 343) 9677 (7311, 12 616) Neurologic 6164 (4301, 8745) Multiple organ systems Neoplastic Total

4444 (2939, 6738) 65 530

Work-Related Conditions

Conditions Awarded SSDI

7806 (4163, 12 229) 2862 (781, 6505) 225 (5, 849) 438 (53,1034) 825 (19, 3106) 1601 (194, 3785) 391 (91, 1957) 1044 (261, 2870) 387 (10, 1935) 387 (10, 1935) 0 (0, 431) 123 (31, 616) 444 (178, 933) 11 843

444 (178, 933)

5134

aExcludes digestive, genitourinary, hematologic, skin, endocrine, and multiple organ systems as there were no work-related cases in the sample. bBased on 71 680 projected SSDI applicants in 1990 (excluding children, psychiatric patients, and widows), and on statewide distribution by organ system. Only organ systems with work-related disabilities in the sample were projected. CI = confidence interval. Cl for projected applicants are based on a standard approximation formula in Hays.28(P29) Projected numbers in the other two columns use exact binomial

calculations.29(PPw.84)

applicants was estimated by organ system, based on the percentage of applications submitted to the Harrisburg office. This distribution was derived from a random sample of cases evaluated in that office, compiled by L. Goldstein during a 1990 quality assurance evaluation. Based on the results of this analysis, the results were extrapolated to Pennsylvania to estimate the annual number of new disability cases statewide that were occupationally related. To do this, an estimate was obtained of the total number of disability insurance applicants for 1990, excluding disabled children, psychiatric cases, and widows. It was necessary to use an estimate because, as of October 1993, no data were available that excluded these categories of applicants from the total. The number of estimated applicant cases in each organ system category was then multiplied by the probability of these cases being occupationally related and the subjects disabled, as derived from the study sample. Finally, the results from each organ system category were summed to generate a total number of estimated occupationally related cases of disability receiving disability insurance benefits in Pennsylvania in 1990.

Results Of the 240 disability insurance appli(69%) were awarded disability benefits and 74 (31%) were denied them. A total of 27 applicants (11%) had work-related conditions, including 8% of those who received benefits and 18% of those whose claims were denied. Table 1 lists the work-related

cants reviewed, 166

cases.

Table 2 shows the distribution of cases by organ system. The distribution in our study sample differed from the overall profile seen in the Harrisburg office; this is because cases are not randomly assigned to claims examiners but are instead

assigned to those examiners with particular expertise. For example, almost all psychiatric claims are reviewed by a psychologist, so these do not appear in the study sample. Our sample had two- to threefold deficits of musculoskeletal (P < .001) and respiratory disease cases (P .021), a twofold excess of multiple organ system cases (P < .0001), and a fourfold excess of cancer cases (P < .0001) compared with the quality assurance sample drawn from the entire office caseload (data not shown). For this reason, results on disability and work relatedness are stratified by organ system. =

ture,2126

even

if other factors may have

contributed to the condition (e.g., lung cancer in a welder). The medical records of applicants were available to the reviewer. However, as others have noted,27 in only a few cases 1788 American Journal of Public Health

did those records yield additional information on the possible work relatedness of the condition. As the probability of a condition being occupationally related varies by organ system involved, the number of


Work Disability

As shown in Table 2, the proportion of applicants receiving disability insurance benefits ranges from 0% (skin) to 86.5% (multiple organ system), and the proportion of these found to be work related ranges from 0% in several categories to 100% in the neoplastic and neurologic categories (based on six cases and one case, respectively). The distribution of claims by organ system in the Harrisburg office may differ from that in the entire state. For example, a small number of claimants with coal workers' pneumoconiosis are reviewed mainly in the Bureau of Disability Determination's Wilkes-Barre branch and are generally processed through a separate, federally funded black lung program. Nevertheless, the distribution of claims by organ system is similar for all three branches of the bureau (H. Thorkelson, Bureau of Disability Determination, personal communication, 1991). The age distribution of our claims reviewer sample did not differ from that of the office-based quality assurance sample (data not shown), and the proportion of cases that were occupational did not vary with age. The results of our extrapolation to Pennsylvania appear in Table 3. We estimated that there would be 71 680 disability insurance applicants in 1990, excluding disabled children, psychiatric patients, and widows (H. Thorkelson, personal communication). Table 3 shows the expected organ system distribution of these cases, based on the distribution in the Harrisburg office (from the quality assurance sample), and the number in each category expected to be both disabled and work related, based on our results in Table 2. The total number of new disability insurance beneficiaries in 1990 with an occupationally related disability was projected to be 5134. Although we did not prospectively abstract information on whether applicants had applied for workers' compensation benefits, we were retroactively able to obtain information on seven applicants. Case 17 was receiving black lung benefits; case 25 had a workers' compensation benefits hearing scheduled; and cases 1, 2, 9, 11, and 13 had no evidence on their disability insurance application of having filed for workers' compensation benefits.

Discussion The study results indicate that a substantial number of applicants for Social Security disability insurance benefits November 1994, Vol. 84, No. I I

in Pennsylvania, and by inference nationally, are likely to be suffering from an impairment caused or exacerbated by prior workplace exposures. Musculoskeletal cases were the most prevalent in our sample despite being underrepresented (based on the quality assurance sample used for comparison). In fact, 40% of the 27 cases judged to have a work-related component were musculoskeletal in origin, and 30% of the 37 applicants in this category had a work-related component. The neoplastic disease category merits specific discussion. Of the 59 disability applicants with cancer, we concluded that, more likely than not, 10.2% had some work-related etiological component. This proportion is consistent with estimates published elsewhere of the annual incidence of work-related cancers.25,30'31 Our finding of 10.2% is probably conservatively biased because information on occupation was abstracted from a disability application form that does not explicitly elicit a full occupational history. With more complete information, more associations between past employment and current cancer might emerge.25'27 Our analysis was based on a small convenience sample from one reviewer's files in one branch of the Bureau of Disability Determination. Consequently, the organ system distribution in our sample differed from that of the bureau's applicant population. We therefore standardized our results to the organ system distribution of the bureau office. This procedure decreased the precision of our point estimates, as reflected in our relatively broad confidence intervals (see Table 3). However, the resulting point estimates are unlikely to be biased, and the statewide projections of work-related impairment and disability reflect the experience of the bureau's applicant population rather than of our sample. The other major source of potential bias in our sample was age. However, the age distributions of our sample and of the

quality assurance sample were similar, making it unlikely that age affected our projected estimates of work-related disease and injury. To date, very few analyses of workrelated disease in relation to disability insurance claims have been made. The only other report assessing work-related disease was a study of asthmatics who completed a Social Security Administration Survey of Disability and Work to assess occupational asthma.32 In that study, unlike ours, the sampling frame included disability insurance applicants

who were denied benefits. Of the 4041 subjects with histories of any active workforce participation, 53 (1.3%) were identified with occupationally attributed asthma. This represented 17.1% of all asthmatics in the sample. Although the diagnosis of asthma was confirmed by physician review in all these cases, the assessment of work relatedness was self-reported for purposes of the study. Limited use has been made of Social Security Disability Insurance data for surveillance purposes. The National Institute for Occupational Safety and Health published data in 1986 on workers who received disability insurance benefit awards or allowances for disabilities in 1975 and 1976.33 Although age-adjusted proportional mortality ratios were disproportionately elevated in many occupational categories for one or more sex and racial groups, the data file used did not include information on the work relatedness of conditions of applicants who were denied benefits. Currently, state agencies responsible for conducting insurance disability determinations do not review applications with an eye toward identifying as sentinel events those claimants who may be suffering from an occupational disease or injury. But because these agencies receive a thorough medical history and at least a sketchy work history from each applicant, they are well positioned to identify claimants (and their coworkers) with potentially work-related impairments. Thus, these disability determination agencies could set in motion various medical and preventive follow-up activities directed at both the claimant and the claimant's former employer(s). These activities could potentially benefit the claimant, other similarly exposed workers, and the Social Security Trust Fund. Disability applicants with impairments that may be occupational in origin might be advised of their eligibility for workers' compensation; since the benefits would generally be more generous, many applicants would likely pursue that route. Follow-up would be appropriate regardless of whether the claimant's application ultimately satisfies the Social Security Disability Insurance test. As currently designed, however, the disability insurance program cannot serve as a freestanding surveillance system for workrelated disorders. At best, program data may add to the aggregate detection of occupationally related impairment and disability. O American Journal of Public Health 1789

Bresnitz et al.

Acknowledgments This study was funded in part by the Preventive Pulmonary Academic Award, No. 1 K07 HLO2100, and by a contract with the Pennsylvania Department of Health. Basic data for use in this study were supplied by numerous sources, including the Pennsylvania Department of Health in Harrisburg, Pa. This paper was originally presented on September 23, 1992, at the Ninth International Symposium on Epidemiology in Occupational Health, Cincinnati, Ohio. The authors wish to thank Mike Ross, Martin Needleman, Dawn Glace, Ed Gracely, and Roseann Bilardo for assistance in preparing this report.

References 1. Preventing Illness and Injury in the Workplace. Washington, DC: US Congress, Office of Technology Assessment; April 1985. OTA-H-256. 2. Baker EL, Honchar PA, Fine LI. Surveillance of occupational illness and injury: concepts and content. Am J Public Health. 1989;79(suppl):9-1 1. 3. National Academy of Sciences. Counting Injuries and Illnesses in the Workplace: Proposals for a Better System. Washington, DC: National Academy Press; 1987. 4. Froines JR, Dellenbaugh CA, Wegman DH. Occupational health surveillance: a means to identify work-related risks. Am J

5.

6.

7.

8.

Public Health. 1986;76:1089-1096. Schumacher MC. Comparison of occupation and industry information from death certificates and interviews. Am J Public Health. 1986;76:635-637. Seligman PJ, Halperin WE, Mullan RJ. Occupational lead poisoning in Ohio: surveillance using workers' compensation data. Am J Public Health. 1986;76:12991302. Centers for Disease Control. Use of workers' compensation claims for surveillance of work-related illness-New Hampshire, January 1986-March 1987. MMWR. 1987; 36:713-714,719-720. Silverstein M, Maizlish N, Park R, et al. Mortality among workers exposed to coal tar pitch volatiles and welding emissions: an exercise in epidemiologic triage. Am J Public Health, 1985;75:1283-1287.

1790 American Journal of Public Health

9. Dubrow R, Sestito JP, Lalich NR, et al. Death certificate-based occupational mortality surveillance in the United States. Am JIndMed. 1987;11:329-342. 10. Fontus HM, Levy BS. Physician-based surveillance of occupational disease: developing a methodology.J Occup Med. 1987;29: 688-691. 11. SENSOR: Sentinel Event Notification System for Occupational Risks. A ProposaL Atlanta, Ga: National Institute for Occupational Safety and Health; 1987. 12. Rosenman KD. Use of hospital discharge data in the surveillance of occupational disease.AmJInd Med. 1988;13:281-289. 13. Schwartz E, Landrigan P. Use of court records for supplementing occupational disease surveillance. Am J Public Health. 1987;77:1457-1458. 14. Maizlish N, Rudolph L, Sutton P, et al. Elevated blood lead in California adults, 1987: results of a statewide surveillance program based on laboratory reports. Am J Public Health. 1990;80:931-934. 15. Blanc PD, Olson KR. Occupationally related illness reported to a regional poison control center.AmJPublic Health. 1986;76: 1303-1307. 16. Bresnitz EA. A model system for occupational disease surveillance activities applying centers for disease control guidelines to poison control centers. J Occup Med.

1990;32:255-259. 17. Rutstein DD, Mullan RJ, Frazier TM, et al. Sentinel health events (occupational): a basis for physician recognition and public health surveillance. Am J Public Health.

1983;73:1054-1062. 18. An Interim Report to Congress on Occupational Diseases. Washington, DC: US Dept of Labor; 1980. Report 60. 19. Social Security Administration. Disability Evaluation under Social Security. Washington, DC: US Dept of Health and Human Services; 1986. Social Security Administration publication 05-10089. 20. Bresnitz EA, Frumkin H, Hodgson M, et al. Occupational Disease and Injury in Pennsylvania. Report to the Pennsylvania Department of Health, Harrisburg, Pa. June 1990. 21. Sferlazza SJ, Beckett WS. The respiratory

health of welders. Am Rev Respir Dis. 1991;143:1134-1148. 22. Lewis R. Plastics. In: LaDou J, ed. Occupational Medicine. Norwalk, Conn: Appleton & Lange; 1990. 23. Selikoff IJ, Lillis R, Nicholson WJ. Asbestos disease in United States shipyards. Ann N YAcad Sci. 1979;330:295-311. 24. Coggon D, Pannett B, Osmond C, et al. A survey of cancer and occupation in young and middle aged men: I. cancers of the respiratory tract. BrJlnd Med. 1986;43:332338. 25. McDiarmid MA. Occupational exposure in lung cancer patients: contribution of remote pastwork.AmJPrevMed. 1991;7:348351. 26. Welch L, Michaels D, Zoloth S. Asbestosrelated disease among sheetmetal workers. Preliminary results of the National Sheet Metal Worker Asbestos Disease Screening Program. Ann N YAcad Sci. 1991;643:287295. 27. Brancati FL, Hodgson MJ, Karpf M. Occupational exposures and diseases among medical inpatients. Prevalence, association and recognition. J Occup Med. 1993;35:161-165. 28. Hays WL. Statistics for Psychologists. New York, NY: Holt, Rinehart & Winston; 1963. 29. Pearson ES, Hartley HO. Biometrika Tables for Statisticians. Cambridge, England: Cambridge University Press; 1966. 30. Markowitz SB, Fischer E, Fahs MC, et al. Occupational disease in New York State: a comprehensive examination. Am J Ind Med. 1989;16:417-435. 31. Balmes J, Rempel D, Alexander M, et al. Hospital records as a data source for occupational disease surveillance: a feasibility study.Am JInd Med. 1992;21:341-351. 32. Blanc P. Occupational asthma in a national disability survey. Chest. 1987;92:613-617. 33. National Institute for Occupational Safety and Health. Occupational Characteristics of Disabled Workers 1975-1976. Analysis of Social Security Disability Benefit Allowances to Workers during 1975-1976. Cincinnati, Ohio: US Dept of Health and Human Services; 1986. DHHS (NIOSH) publication 86-106.
