Surveillance of Occupational Illness and Injury in the United States: Current Perspectives and Future Directions Author(s): Edward L. Baker, James M. Melius and J. Donald Millar Source: Journal of Public Health Policy, Vol. 9, No. 2 (Summer, 1988), pp. 198-221 Published by: Palgrave Macmillan Journals Stable URL: http://www.jstor.org/stable/3343005 Accessed: 05-12-2015 18:13 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Journal of Public Health Policy.

http://www.jstor.org

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

of OccupationalIllnessand Injury Surveillance in the UnitedStates:CurrentPerspectives and FutureDirections EDWARD L. BAKER, JAMES M. MELIUS,

and J. DONALD MILLAR

INTRODUCTION: FOR

TRENDS

OCCUPATIONAL

IN SURVEILLANCE DISEASE

AND

PROGRAMS INJURY

ODERN systems for the surveillanceof disease and injury have three components:collection of data, analysisof data, and "thedisseminationto all who need to know" (i). Inherentin this formulationdeveloped by Dr. Alexander Langmuir,surveillanceis being perksc9eA formed so that "correctivemeasureswill be taken." In the recent past, surveillanceof occupational disease and injuryhas focused primarilyon developing techniques for gatheringand analyzing data with relativelylittle attentiongiven to disseminationand response. Futureefforts at surveillanceof occupationaldisease and injuryshould be motivatedby the desireto use data in a way that will lead directlyto active interventionfor the preventionof these conditions. Unfortunately,in the mindsof manyoccupationalhealthprofessionals, surveillancesystems are viewed as passive and ponderous systems designed to collect informationof uncertainutility.To achieve a broader involvementof occupationalhealthprofessionalsin surveillanceof occupationaldisorders,systemsmust be developedthat areintrinsicallyactive and allow for rapid responseto emergingtrendsof illness and injury. In accomplishingsuch a transition,the ultimate goal is to develop a surveillancesystem that has the capacity to respond to the presenceof workplace hazardsand to provide data that direct the efforts of health professionalsto intervenein the workplace.Furthermore,the usefulness of the surveillancesystem should be immediatelyapparentto occupational health professionalswho contributedata to the system. Unfortunately,many surveillancesystemsfail throughan inabilityto demonstrate I98

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

I99

that dataareusedto directinterventionefforts.At the presenttime,the NationalInstitutefor OccupationalSafetyand Health (NIOSH)is involvedin aneffortto improveexistingsurveillance systemsandto develop new approachesto identifyoccupationalillnessand to monitortrends of occupationaldiseaseand injury. To be successful,surveillanceprogramsrequireinteractionbetween those who collector controlaccessto data and otherswho analyzeit anddisseminatethe results.Clearly,if thesurveillance effortis to be used to influenceinterventionprograms,the systemshouldbe designedwith the full inputof data collectors,data analyzers,and those who would intervenein the workplace. Thispaperis intendedto serveseveralpurposes:first,to reviewconceptual issuesrelatingto surveillanceof occupationaldiseaseand injury; next, to identifykey gaps in currentprograms;and finally,to indicate futuredirections.The conceptspresentedherewill form the basis for NIOSHsurveillanceeffortsin the future. ESSENTIAL OCCUPATIONAL

PURPOSES HEALTH

AND AND

CHARACTERISTICS SAFETY

OF

SURVEILLANCE

As recentlydefined,"Epidemiological surveillanceis the ongoing,systematiccollection,analysis,andinterpretation of healthdataessentialto the planning,implementation, and evaluationof publichealthpractice, closely integratedwith the timelydisseminationof these data to those who needto know"(z). Fromthisdefinition,it is evidentthatsurveillance mustbe directlylinkedto preventiveaction. Publichealth surveillancesystemsmust have "afferent,""central," and "efferent" componentsto achievefunctionalintegrityandto address the publicneed.Inputcan come frommanysources:physicianreports, vital records,medicalexaminerfindings,or laboratoryreports.Central processingis requiredto determineif the inputis of sufficientqualityto merit reportingand response.In the case of occupationalhealth,the actionspromptedby the surveillance systemshouldbe directednot only at the individualcase or the affectedgroupbut also at the responsible workplace factors. Typically,for occupational health surveillance,one or more components has significantlimitations.Forexample, many states

havelaws requiringphysiciansto reportselectedoccupationaldiseases; however,since follow-uphas not occurredroutinely,and penaltiesfor nonreporting areabsent,reportsarenot filed.Evenif reportingdidoccur, statehealthdepartmentsare often understaffedin occupationalhealth

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

200

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

Further,mostlacklegalauthorityto and cannotrespondappropriately. to identify hazards.A clearneedexistsfor enterandinspectworkplaces approachto occupationalillnessandinjurysura morecomprehensive veillance. systemshould: In occupationalsafetyandhealth,a surveillance i) identify cases of occupationalillness or injury,and/or z) monitor trendsof occupationalillness or injury.

IdentifyingCases In occupationalhealth (as in communicablediseaseprevention)(3), the purpose of case identificationis to target an interventionof directvalue to the affected individualand to others at risk of developingthe same disorder.In effortsto controlcommunicabledisease,e.g., smallpoxeradication or malariacontrol (4), case identificationis followed by attempts to identifyother individualsat risk and to control environmentalfactors that areresponsiblefor diseaseetiology.In occupationalhealth,relatively little has been done to develop a system that identifiescases and directs appropriatepublic health practice.A comprehensiveprogram for case identificationof occupationaldisorders("SENSOR")is describedbriefly in this paper. Monitoring Trends Surveillancein occupationalhealthhas traditionallyfocusedon monitoring trends of illness, injury,or exposure to workplace hazards.In such activity,surveillancedata are developed to assess variationsin rates between i) differentindustrialgroups, z) differentgeographicareas,and/or 3) differenttime periods. From such comparisons, health officials can identify target industriesor geographicareas requiringfurtherresearch or intervention.Ideally,trendmonitoringshould be used to evaluatethe efficacy of specificprogramsdesigned to control occupationalhazards. This manuscriptdescribessources of data that can be used to monitor trendsand discussestheir advantagesand disadvantages. SURVEILLANCE

TARGETS

Surveillanceof occupational disease and injurycan involve monitoring of either health conditions in the workforce, or hazardspresent in the workplace. Although health effects surveillanceand hazardsurveillance can be performed as separate, distinct endeavors, linkage of the data derivedfrom health effectsand hazardsurveillancefor the same popula-

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. - OCCUPATIONAL

ILLNESS

AND

INJURY

zoi

TABLE I

TenLeadingWork-Related DiseasesandInjuries i. Occupational LungDiseases z. Musculoskeletal Injuries 3. OccupationalCancers(otherthanlung) 4. SevereOccupationalTraumaticInjuries

5. Occupational CardiovascularDiseases

6. Disordersof Reproduction 7. NeurotoxicDisorders 8. Noise-InducedHearingLoss 9. DermatologicalConditions io. PsychologicalDisorders

tion is often preferable. In certain situations, isolated surveillance of individual health effects or workplace hazards may be useful.

HealthEffectsSurveillance To focus prevention and research efforts, NIOSH has identified Ten Leading Work-RelatedDiseases and Injuries (5) (Table i), and has developed national strategiesfor the preventionof those conditions. In this group of disorders, there are a few which are uniquely attributableto workplace etiologies. These conditions, referredto by the World Health Organization (WVO) (6), as "occupational diseases and injuries," are caused by a single exposure or hazard (e.g. silicosis, lead poisoning, and noise-induced hearing loss). In view of the relative clarity of cause and effect relationships, surveillanceof such conditions is conceptually simple. Since surveillancein occupational health is in an early stage of development, systems should focus on those conditions most directly attributableto work. Certain conditions on the NIOSH list are of multifactorialetiology, with one of the factors being relatedto work. Health effects surveillance for multifactorialdiseases is more complex and must involve the ascertainment of all factors, both work-related and non-work-related,suspected of playing a role in the disease process. In some cases, e.g. hazard surveillanceas discussedbelow, surveillanceof the responsibleworkplace exposure or hazard may be more useful than health effects surveillance in targetingpreventionprograms.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

202

JOURNAL

OF PUBLIC

HEALTH

POLICY

SUMMER

I988

Data Sources A wide varietyof sourcesof data exist for potential use in health effects surveillance.Frequently,health status is assessedby directevaluationsof individuals (i.e., surveys)using a questionnaireand medical diagnostic tests. These surveysmay be directedat a specificoccupationalcondition (e.g., asbestos-relatedsource),the health effectsof work generally,or the generalhealth of the individualworker.When individualevaluationsare performed,ascertainmentof some measureof exposure (i.e., job history and/or biologicalor environmentalmonitoringresults)is often combined with the assessmentof health status. Healtheffectssurveillancecan also be accomplishedby analyzingexisting medical records (e.g., hospital dischargerecords,death certificates, or birth records). By coding information on industry and occupation from such records,surveillancecan be performedfor trendsin the occurrenceof healthconditionswithin occupationand industrygroups.Other recordscontaining medical diagnoses are currentlybeing generatedfor purposesnot directlyrelatedto the provisionof healthcare.Forexample, workers'compensationclaims contain diagnosticinformationas do employer-generatedrecords required by regulators of the Occupational Safety and Health Administration(OSHA). These data sources are discussed below.

HazardSurveillance Hazard surveillanceconsists of periodiccharacterizationof chemicalor physical hazardsin the workplace. Since many chemical, physical, and biological agents have been found to presenta significanthazardto the health of workers, surveillanceof the occurrencesof hazardsprovides very usefulinformationeven in the absenceof simultaneoushealthstatus assessment (8-iO). For example, the National OccupationalExposure Survey(NOES) representsa unique large-scalesystem of surveillanceof occupationalhazardsin a sampleof U.S.workplaces(i i). In manyindustries, hazard surveillance(i.e., direct measurementof levels of airborne contaminantsor noise level) is used to directstrategiesfor primaryprevention. LIMITATIONS COMMUNICABLE

AND

APPLICABILITY

DISEASE

OF THE

SURVEILLANCE

MODEL

Much of the nation's public health perspectiveon surveillanceis based on the extensiveexperienceof the Centersfor Disease Control and state

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

203

and local healthdepartmentsin surveillanceof communicablediseases (3) such as polio, measles,hepatitis,and influenza.Althoughcertain aspectsof the communicablediseasesurveillancemodelare applicable for surveillanceof occupationalconditions,thereare basic differences betweencommunicablediseasesand occupationaldisorderswhich impact on the strategyfor surveillance: i) Strongdisincentives exist for reportingoccupationalconditionsby employers,physicians,and, most importantly,affectedindividuals. Suchdisincentives communicable do not characterize diseasesurveillance. z) Healthprofessionalslack knowledgeregardingthe natureof health risksin the workplace.Generally,they aremorefamiliarwith communicablediseasesurveillance. 3) Manycommonoccupationalconditionsarealsocausedoraggravated by non-occupational factorsunrelatedto work. 4) The latencyperiodbetweenexposureonsetanddiseaseoccurrenceis frequentlygreatfor occupationaldisorders. Despitethesedifferences,significantsimilaritiesalso existthatallowfor a parallelismbetweensurveillancesystemsfor occupationalillnessand injuryand surveillancefor communicabledisease. i) They both exist for the samereason-to facilitatepreventionof disease and injury. of reportingproceduresis essentialfor the developz) Standardization mentof meaningfuldatafromeithersystem. of stateandlocalpublichealthagenciesandhealthcare 3) Involvement providersis typicallyessentialfor the generationof timelyandaccurateinformation. 4) Surveillanceis an acknowledgedfunctionof publichealthagencies andis perceivedas an appropriate meansof discharging thestatutory responsibilities of theseorganizations. CONTRASTS ILLNESS

BETWEEN

INJURY

AND

SURVEILLANCE

The needsof occupationaldiseaseand injurysurveillancearesimilarin manyrespects.Bothneedsensitiveandspecificsurveillance systemsthat providemeaningfulresultssuchthat appropriateinterventionstrategies can be developedand implemented.To a certainextent,surveillanceof injuriescanemploymethodsanddatasetsthathavebeenusedto conduct illnesssurveillance.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

204

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

However,the significantdifferencesthat exist presentuniquechalManyocfor occupationalinjurysurveillance. lengesandopportunities inetiology. cupationalillnessesmaybedifficultto identifyasoccupational Attributingtheseillnessesto a specificworkplaceexposuremaynot be illnessesmayhavelonglatencyperiodsand possible.Manyoccupational however, effortsdifficult.Injuries, multiplecausesthatrendersurveillance in mostcases.Definingthe canbe easilyidentifiedas beingwork-related injuryusuallyrequiresa knowledgeof theimmediatesequenceof events occurringa few minutesto a fewhoursbeforetheinjury.Reconstruction thanin thecaseof occuof thecausalpathwayis muchlessproblematic pationalillness. traumadisorders(CTD)(insomewaysbest of cumulative Surveillance describedas "chronicinjuries")sharescommonrequisiteswith both (I 3). Sincephysicalfactorsareinvolvedin injuryandillnesssurveillance the etiologyof theseconditions,theircausesare somewhatsimilarto thoseof acutetraumaticinjuries.Thesedisordersmayactuallyrepresent minoracuteinjuries.SinceCTDmay impactof repetitive, thecumulative occuraftera latencyperiodof weeksto monthsandsincemultiplefactors programsfor CTD surveillance may be responsiblefor thisoccurrence, illnesses. willsharemanyfeatureswiththosedevelopedforoccupational CURRENT

DATA

SOURCES

SURVEILLANCE-USEFULNESS

FOR

HEALTH AND

EFFECTS

LIMITATIONS

HealthCareProviderRecords Manytypesof healthrecordscontainphysiciandiagnosesthatmayrep(includDeathcertificates forsurveillance. resentconditionsappropriate hospitaldischargerecords, ing thoseon fetaldeaths),birthcertificates, officerecordsof primarycareproviders(includingemergencyrooms), insuranceclaim files, and recordsof patient visits to occupational medicineclinicsrepresentpotentialdata sourcesfor surveillanceactivities.Theircurrentlimitationsinclude: disorof occupational or misclassification nonrecognition Physicians' ders z) Lackof occupationalinformationon the record 3) Lackof timelinessof reporting to reporting 4) Theinfluenceof disincentives i)

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

205

The advantagesinclude: i) Availabilityat relativelylow cost z) Coding schemes exist (e.g. InternationalClassificationof Diseases)

Despite the limitations of diagnostic records,the process of collecting data on occupational conditions may serveto improveawarenessamong health care providers of the impact of work on health. By educating health care providers in recognizing suspect occupational disorders as part of implementinga surveillancesystem, preventionof disability can be improved through early detection. Furthermore,if by using these methods, employers are motivated to institute environmentalcontrols, future cases may be prevented.

DirectMedicalSurveys OSHA regulationsrequiremedical evaluationsfor a varietyof conditions (e.g., asbestosis, coal workers' pneumoconiosis, byssinosis, and lead poisoning). More broadly, employers perform periodic medical evaluations to assess the individual health status of employees. Such periodic evaluationsmay be performedas part of a largereffort to promote health (i.e., "wellness"programs)or to monitor workers for health risksrelated to their work. In some situations, medical evaluations are performed before beginninga potentiallyhazardousjob to provide baselinedata for subsequent medical surveillance, or to identify health conditions that could be exacerbated by specific hazards associated with the job. By evaluatingindividual workers using medical questionnaires,diagnostic tests, and physical examinations, health surveillancecan be performed directly.The major limitations include: i)

High cost

z) Nonstandardizationof examination techniques 3) Inadequate characterizationof exposure to relevant chemical and physical job hazards The important advantagesare: i) Examinations can be designed to evaluate health effects of specific exposures or physical hazards. z) Repeated testing of the same individualprovides longitudinal assessment of the effects of changing exposure on health. 3) Detection of early or subclinicalabnormalitiescould lead to a direct interventionbenefitingthe affected individual.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

zo6

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

4) By measuringexposurestatusand healthstatussimultaneously,exposure-responserelationshipscan be developed. 5) Workersare more aware of the health issues surroundingtheirwork as a resultof participatingin a health surveillanceprogram. 6) Accurate,generalizablerates can be derivedfrom properly-designed surveysof definedpopulations. To providemaximumvalue,data from directsurveysshould be collected in a standardizedfashion and pooled in a mannerthat will allow direct retrievaland analysis. Employer-GeneratedRecords In responseto OSHA regulations,employersmustmaintaina log (OSHA Form zoo) of injuriesand illnesses in the workplace that conforms to criteriadefinedby the Departmentof Labor.Currently,such recordsare severelylimitedas sourcesof surveillancedata. Employershave financial disincentives(e.g., workers'compensationclaims)to reportillnessesand injuries.Furthermore,the duty of compilingthese recordsoften falls to an individualwithout trainingin the accuraterecognitionof occupational disorders.Thus, accuracymay not be a high priority.Priorstudiesshow that the OSHA Form zoo log may only recorda fraction of the illness and injuriesin the workplace (I4). Workers'CompensationData Eachstatethat maintainsa workers'compensationsystemgeneratesdata of potentialuse in surveillance.Unfortunately,for data to be enteredinto the system, a worker must: i) recognizehis / her condition as work-related, and z) file a claim. To receivecompensation,the workermust also i) satisfystateregulationsfor eligibility,and z) successfullywin a decision by the workers'compensationboard. Limitationsof workers'compensationdata for surveillanceare: i)

In view of reportingdisincentivesand inherentdifficultiesin recognizing occupationaldisorders,workers'compensationdata will consistently underestimatethe true rate of occurrenceof occupationaldisorders. Furthermore,the rate of underestimationwill vary between conditions,with greaterunderreportingfor diseasesthan for occupational injuries.

z) Many workers' compensation systems have significant limitations (e.g. requirementsthat claimsbe filedwithin a fixed periodfollowing

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

207

the suspectexposure)whichprecludefilingclaimsfor manyoccupationaldiseases(e.g.cancer). 3) Obtainingmeaningfulnationaldata is difficultsince reportingrequirementsvarybetweenstates. Nevertheless,significantadvantagesto theuseof workers'compensation datainclude: i)

z) 3) 4) 5)

6)

All recordsin the data set relateto conditionsof suspectedoccupationaletiology. Informationon the job and industryfor each claimis containedin the record. The circumstances of the illnessor injuryarefrequentlydescribedin a way thatprovidesunderstanding of the causeof the condition. If case identificationleadsto improvement of workplaceconditions, preventionof furtherclaimsshouldoccur,thus benefitingboth employeeand employer. If these data are used for surveillancepurposes,technicalimprovementsin the datamanagementsystem(e.g.,bettercodingprocedures or computersystems)would occurthat would benefitthe managementof the workers'compensationinsurancesystemitself. Finally,the recordsarecurrentlyavailable.

In summary,workers'compensationdatarepresentan importantsource of surveillance datathatcan be usedto monitortrendsin the occurrence of selectedoccupationaldisorders(I 5) andto identifycasesforfollow-up action.

LaboratoryData Certaincommerciallaboratoriesanalyzebiologicalsamples(e.g.,blood and urine)to determinethe concentrationof toxic substancesabsorbed by workersfromtheirenvironment.Laboratorytestresultscan be used as anindexof diseasestatusin selectedinstances.Thisapproachis limited by severalfactors: AlthoughfirmsarerequiredbyOSHAregulations(e.g.,leadstandard) to performtests,somedo not. Smallfirmsmaycomplylesscompletely thanlargefirms. z) Workerswithin a companyare usuallynot requiredto submitto testing;therefore,an incompletesampleis obtained.Workerswho i)

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

Z08

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

of an abnormaltestresultmayselectivelyrefuse feartheimplications to participate. biologicalassaysexist for only a veryfew sub3) Most importantly, stances encounteredin the workplace.For those substancesthat can be analyzedin body fluids,laboratoryquality-controlprogramsexist for only a few. Despite these limitations,clinical laboratorytests have importantadvantages: Each test (e.g., blood lead concentration)is specificfor the diseaseof interest.Furthermore,the severityof toxicity is usuallyproportional to the blood or urine concentration(I6). z) In some states (e.g., New York and Maryland), commercial laboratoriesperformingsuch tests are requiredto reportresultsto a centralstate agency office for statisticalanalysis. i)

In summary,although the use of laboratory-basedsurveillancemay be limited to a few agents, it would include agents, such as lead, that may impact a large number of workers. This type of surveillance,used in conjunctionwith other surveillanceactivities,may serve as an efficient and relativelyinexpensivesystem for identifyinghigh risk worksitesand suspectedcases of occupationaldisease. Summary

Data bases that can be used for occupationaldisease and injurysurveillance are many; each has inherentstrengthsand weaknesses.When the decision is made to place a condition undersurveillance,a strategymust be developed to evaluate existing potential data bases or identify new data that must be collectedfor effectivelytrackingthe condition.To help set prioritiesfor developingdata sets, NIOSH personnelhave reviewed existing data sourcesand evaluatedtheirusefulnessin case identification (Tablez). In general, the identificationof cases is best accomplishedthrough activities closely related to the practice of clinical medicine. Medical surveys typically consist of questionnaires,physical examinations, and tests which evaluate organ system function- a format similar to the typical content of a physicianvisit. From such visits recordsare derived to facilitateongoing patient care or to provide for payment of medical costs. In monitoringtrends,the usefulnessof these data sourcesis similar, but not identicalto that noted in Table z.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

O0

z

-Z

Z

I~~~~~~~~~~~' v~ 4

* 4

I

'W

'I,

4

C)

S-4~~~~~

S-4~~~~~~ 0~~~~~~~~~~~~~~~~~

ZZ UU~~~~~~~~~~~~~~~~~ .uO

H 0 ot 0

0 =-wt t H

o

U ...4

I,

~

~ ~ ~Q~~~~~~~~~~C ~ ~ ~ CZ ~~~~~~~~~~~

0 ~~ ~ '4.~~~~~~~~~~~~~ Ct

H

~

0

4U

U4

0~~~~~~~~~~~~~~~~~~~~~~~Z O

N

;G;)0

O

U

"I

0 4:

S-

U

0

-~~~~~~~~~~~~~~~-

~

H

H

u1

~

tE

~

U

.

-I

iArn-

S-~~sz -:

a

u 4,1 11

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

2IO

JOURNAL

OF PUBLIC

HEALTH

POLICY

- SUMMER

1988

In most situations,diagnosticrecords,if improvedin quality,and direct health surveyswill provide the most useful health data for surveillance purposes.The initialpurposefor collectingdata should be scrutinizedin consideringtheiruse in the surveillanceof occupationalillnessand injury. In view of the complex issues affectingthe qualityof statisticsfor occupational illness and injury,designingsystemsspecificallyto collect information for surveillance,rather than adapting existing systems created for other purposes, is generallypreferableeven though it may be more expensive. In certain cases, direct use of existing data (e.g., laboratory test results)will be useful for surveillanceof specifichealth effects. CURRENT

DATA

SOURCES

FOR

HAZARD

SURVEILLANCE

OSHA InspectionRecords Information from inspections performed by OSHA inspectors has recently been proposed as a source of hazard surveillancedata (9). Such data are useful becauseactualenvironmentalmeasurementsare made in a largenumberof worksites,yieldingindustry-specific hazardrates.These data are compiled following direct inspection by trained industrial hygienists.Severallimitationsexist for this type of data: Recordsexist only for substancesregulatedby OSHA and constitute a rather small percentageof the large number of potentially toxic substancesin the workplace. z) OSHA inspections are targetedin variousways that may yield data not representativeof hazardoccurrence. 3) Results reflectconditions only on the day of inspectionand may not be typical of day-to-dayplant conditions. OSHA inspections do not include small businesses, government 4) facilities,and other workplaces. i)

Advantagesinclude: Trained inspectors make direct measurementsof level of airborne contaminantsusing standardmethods. z) Data are coded in a way that can generate estimates of industryspecifichazardrates. 3) Data collection procedureshave not varied significantlyover recent years or between differentgeographicareasof the United States. i)

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

2II

Direct Hazard Survey Hazard information exists for a large arrayof U.S. industriesas a result of two NIOSH surveys statisticallydesigned to provide a representative sample of U.S. workplaces: the National Occupational Hazard Survey (NOHS), and the National OccupationalExposureSurvey(NOES) (i i). Identifying chemical constituents of trade name products has further refinedthe data from these surveysand increasedtheirvalue. Limitations include: I) Direct surveys are very costly. z) Developing a final report requiresyears of effort because of the large amount of data collected. 3) Since most workplace substances are identified only by their trade names, significant effort must be expended to identify constituents by their generic chemical names. 4) In view of the scale and complexity of this type of survey,collecting industrialhygiene samples to characterizeworkplace concentrations of specific airborne constituents is not feasible.

Advantagesare: Since the survey sample can be drawn by statisticalprocedures, national estimates can be derived from survey results. z) In addition to providingestimatesof hazardor exposure ratesby type of workplace, the survey develops a file of trade name ingredients that contains the chemicalconstituentsof a largenumberof industrial products. i)

In summary, direct surveys, although expensive and time-consuming, provide the best estimates of the occurrenceof hazards or exposures in the workplace. These data are particularlyuseful in monitoring secular trends between industries and for different time periods. Currently, NIOSH is conducting an exposure survey in the mining industry. AN

OVERVIEW ACTIVITIES

OF CURRENT AND

SURVEILLANCE

OPPORTUNITIES

State Health Department Programs As interest in occupational health problems has increased within communities across the country (I 7), state departmentsof health and labor

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

2I2

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

TABLE 3

Activities Surveillance Occupational StateHealthDepartment State

California Colorado Iowa Kentucky Maine Massachusetts Nebraska NewJersey NewYork

NorthCarolina NorthDakota Ohio Pennsylvania RhodeIsland SouthCarolina Utah Virginia Wisconsin

Program

-Lead Poisoning Surveillance Laboratory Diseases of Occupational Reporting Physician InjurySurveillance Work-related Disorders forOccupational Surveillance HospitalDischarge for Surveillance DeathCertificate SentinelHealthEffects Occupational of Occupational Diseases Reporting Physician HazardSurveillance Reproductive CancerSurveillance Occupational BirthDefectsSurveillance IllnessinFarmWorkers Pesticide-related SilicosisSurveillance Disorders VitalRecordSurveillance forOccupational HeavyMetals Laboratory Surveillance, LungDiseaseRegistry Occupational for Surveillance DeathCertificate Disorders Occupational InjurySurveillance Occupational for DeathCertificate Surveillance Disorders Occupational CancerSurveillance Occupational Disorders forOccupational Surveillance HospitalDischarge Surveillance DeathCertificate forOccupational Diseases DeathCertificate Surveillance HazardSurveillance Reproductive of Occupational Diseases ActiveSurveillance of Pneumoconiosis ActiveSurveillance

haveexpandedtheiractivitiesin monitoringandcontrollingoccupational hazards. Most have receiveddirect NIOSH support (Table 3). A need exists in all these organizationsto expand their involvementin occupational diseaseand injury;in some instances,specificsurveillanceprojects are now under way.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

2I3

Severalstatesaredevelopingsurveillancesystemstargetedspecifically at occupationalinjuries.For example,Coloradooccupationalhealth professionalshave developeda comprehensivesurveillancesystemfor occupationalinjuriesbasedon Workers'Compensationreports,records of the Vital Registry,and Departmentof Highways.They have also developedcomparabledenominatordata from Censustapes enabling themto generateinjuryratesspecificfordemographic characteristics and geographicalareas.UtahandNorthDakotahavealso developedinjury surveillancesystems. The surveillancesystemof the VirginiaStateHealthDepartmentuses workers'compensationrecordsto recordoccupationalinjuriesas well as illnesses.Afterthe EpidemiologyDivisionrecordsthe occurrenceof a condition,interactionwith otherstateprogramshas led to plantvisits designedto evaluatepossiblecausesandto recommendcorrective action. The New JerseyStateHealthDepartmenthas initiateda surveillance systemfor silicosisthathas standardized data-collection proceduresand usesexistinghealthdepartmentresourcesto follow-updiagnosedcases. An essentialfeatureof this surveillanceactivityis the collectionof data thatwill leadto immediateactionin theworkplaceof affectedindividuals.As anotherexample,theWisconsinHealthDepartmentis conducting surveillanceof pneumoconiosisin whichdifferentdatasourcesareused to assesstheprevalenceanddeterminants of diseaseoccurrence. Hospital dischargerecords,physicianreports,and workers'compensationdata arebeingcomparedto evaluatetheirdifferentialuse. CurrentNIOSH Programs CurrentNIOSHsurveillanceactivitiesaredesignedwith one basicgoal in mind:the establishment of nationaland statesurveillance systemsfor conditionsincludedon the NIOSH List of Ten LeadingWork-Related Diseasesand Injuries.As these surveillancesystemsidentifyimportant trendsor the occurrenceof cases of suspectedoccupationaldiseaseor injury,interventionprogramscan be developed.Six specificobjectives havebeendevelopedto focusNIOSHeffortson achievingthis goal: i) Developa modeloccupational diseasereportingsystemlinkinghealth careproviderswith statehealthdepartments z) Incorporateoccupationalhealth issues into nationaland regional surveys(e.g.NationalHealthandNutritionExaminationSurveyand HealthInterviewSurvey)

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

2I4

3) 4) 5) 6)

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

Improvehazard surveillance Develop uniform approachesfor using existing health data sources Disseminateinformationand share methodology Performsurveillancefor the Preventionof the Ten LeadingWork-Related Diseases and Injuries

A wide varietyof surveillanceactivitiesexists within NIOSH (I8,I9). Severalaretrulynationwidein scope and can be usedto generatenational estimatesof the occurrenceof certaintypes of occupationalillnessesand injuriesand to providetrenddata on workplaceexposures.Directlinkage of such surveillanceactivitiesto primarypreventionprogramsoperating at the local level has not yet been achieved to any appreciabledegree. Surveillanceactivitieshave served many useful purposes (e.g., development of the trade name ingredientfile) and have generatedhypotheses suggestingfurtherdirectstudy. Other Activities Over the past decade, the scope of the occupationalhealth servicesprovided has broadenedand becomemoresophisticated.Forexample,occupational health clinics have developed throughout the country to serve the needs of workers and employers by providing evaluationsof individual cases, surveysof exposed individuals,and education of workers regardingworkplacehazards (zo). Occupationalhealth clinics have recentlyformed a national network to standardizeprocedures for data collection and to help the clinics performcollaborativeresearch.Although federaleffortswere originally designedto assist the clinics in the study of the control of occupational cancer, efforts currentlyunder way will also help the clinics perform surveillanceof other categoriesof occupational disease. Recently,New York State has started to develop a statewide network of occupational health clinics across the state. These clinics will focus on the diagnosis, treatment,and preventionof occupationaldiseasesin the state. Large-scale, computerized information systems for occupational health have been installedin the health programsof large U.S. corporations to help corporatehealth personnel to assess the currentstatus of the health of their workforce and to allow for monitoringof long-term trends.Althoughthese systemshave accumulatedlargeamountsof data on exposure and the health status of employees, their potential in epidemiologicsurveillancehas not been realized.Poison control centers

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

2I5

involvedinrespondingto concernsaboutoccuhavebecomeincreasingly pationalhazards(AI); informationgatheredby such centerscould be for use in occupationaldiseasesurveillance. standardized

GAPS

IN

CURRENT

SURVEILLANCE

ACTIVITIES

the Collectionof Surveillance Data Standardizing Manycurrentsystemsof diseasesurveillance relyheavilyon thediagnosis the abilityof of occupationalconditionsby physicians.Unfortunately, physiciansto diagnoseoccupationalillnessandinjuryis inadequate.This is due both to weaknessesin physicianeducationfor the recognitionof occupationalconditions(zz) andto a generalreluctanceon the partof physiciansto becomeinvolvedin thediagnosisof occupationaldisorders. Forexample,evena well-recognized conditionsuchas pulmonaryasbestosishas beendiagnosedusinga widevarietyof criteria.Uniformcriteria for identifyingreportable"suspectedoccupationaldisorders"have adof reportablecommunicable vantagesanalogousto those characteristic diseases(3). Suchreportingcriteriashouldbe designedto meettheneeds of epidemiologicsurveillanceratherthanclinicaldiagnosis.In this context, the key concernis to achieveconsistencyof reportingby different datasources. This need for standardization contrastswith the processof clinical diagnosisthatprovidesforindividualassessmentof eachperson.Reportingcriteriacouldformthebasisforassessingtherelativevalueof different sourcesof health surveillanceinformationin additionto providinga guidefor activesurveillance. Methodsfor ImprovingExposureAssessment Identifyingoccupationalillnessandinjuryreliesnot onlyon characterizingthemanifestations of illness,butalsoon knowledgeof theindividual's job experience.Althoughsystemsexist for codingthe occupationand industryinformationof employedpersons,suchinformationis usually not linkedwithjob-specific chemicalandphysicalhazarddata.Forexample, knowledgethatan individualis employedas a coremakerin a brass factoryis usefulonly if knowledgeaboutthe typicalchemicalexposures in thatworkplace(e.g.,lead,carbonmonoxide,andsilica)is alsoavailable in a systematicway.Developinga job-hazardmatrixwillgreatlyfacilitate

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

2i6

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

the abilityto link specificchemicalor physicalhazardswith diseases occurringin workingpopulations. Centersand ProviderReportingNetwork LocalSurveillance localeffort.FreA successfulsurveillance programrequirescoordinated quentlystateand local activitiesrelatedto occupationalhealthsurveillance are fragmentedand inadequatelysupported.In manystates,no surorganizationexistsat the statelevelto coordinatea comprehensive veillanceof occupationaldisorders. Centers,basedin StategovernA nationalnetworkof Surveillance -would facilitate ment-typically thehealthand/ or labordepartments dataandcoordinateresponsesto the uniformcollectionof surveillance occurrenceof injuriesor illness.An importantfunctionof sucha center wouldbe to linktogetherexistingunrelatedstateprograms,eachhaving of occupational illnessandinjury.Existimportantrolesintheprevention ingprograms,suchas providerreportingof occupationaldisease,could be coordinatedand improved.A uniform approachto providerreporting -a SentinelEventNotificationSystemfor OccupationalRisks (SENSOR)-has been proposed by NIOSH as a state-basedeffort for active surveillanceof occupationalillness and injury(23). SENSORprojectscould use existing data sources to performsurveillance for case identificationand to monitor trends. Case identification through a providerreportingnetwork would permit active surveillance, and, most importantly,directedintervention.To some extent, SENSOR centers would resemble the Surveillance,Evaluation,and End Results (SEER)projectsestablishedby the National CancerInstitutefor cancer surveillanceat the state or regional level. Each state surveillancecenter would transmitdata periodicallyto a centralofficeat NIOSH, andwould receivefrom NIOSH national summariesand data from SENSORprojects in other states. CentralCoordinationof SurveillanceActivities Centralcoordinationis essentialfor maintaininguniformityin data collection between the surveillancecenters and monitoring data quality. NIOSH is uniquelyable to act as a centralcoordinatorfor stateprograms by supportingstate activities,providingtechnicalexpertiseto these programs, organizing national meetings and other forms of information exchange, and facilitating interaction among the states and between states and federalagencies.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL DIRECTIONS

FOR

AND

ILLNESS

FUTURE

INJURY

ZI7

ACTIVITY

Project SurveyStandardization In a manner analogous to prior efforts (24) to standardizethe methodology for pulmonary function evaluation in population surveys, NIOSH has initiated a project designed to develop a standardquestionnairefor obtaining an occupationalhealth history.The questionnairedevelopment process will be similar to that used by the British Medical Research Council and the AmericanThoracic Society (25) which utilizedan expert consensus approach to develop the questionnaire.To help standardize the questionnaireand other health testing methods, NIOSH will use the facilitiesand surveysof the National Centerfor Health Statistics,particularly the National Health and Nutrition ExaminationSurvey(NHANES) and the National Health InterviewSurvey (HIS).

SentinelEventNotificationSystemfor OccupationalRisks(SENSOR) Drawing on existing State laws that mandate the reporting of occupational disease, NIOSH will fund a limited number of projectsto demonstrate the feasibility of SENSOR-an effective surveillancesystem that can detect and intervene in individual cases of occupational disease. Essentialto the success of this effort is the developmentof state projects which will respond to individual cases, perform medical evaluations of coworkers at risk, and investigateworksite hazards.Successof the SENSOR projects will depend on the ability of state agencies to coordinate a unified effort to monitor the occurrence of selected conditions and to develop a targetedpreventionprogram.

HazardSurveillance Surveillanceof workplace hazards representsan important and largely unexplored approach to targeting control of occupational illness and injury.To achievethe potential that exists, data sourcesmust be identified and evaluated,and methods for improvingaccess to exposure data must be developed (e.g., a uniform job-exposure matrix).

SurveillanceUsingExistingHealthData Sources Many health data sources can be useful in monitoring disease and injury trends and occasionally in identifying cases for intervention.Each data source must be examined to assess data quality and value for surveillance

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

2i8

JOURNAL

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

purposes (z6). Since each data sourcewas developedfor purposesother than surveillance,its use for a differentapplicationshould be assessed. For those data sets that are found to be useful, improvementsmay be needed to achieve an appropriatelevel of precision and timeliness. A method has been proposed (z7) that will be expresslyusefulin the objective evaluationof existing data sourcesfor surveillance. National surveysperformedby otherfederalagenciesrepresentimportant existing data sources. As these surveysare revised (e.g., NHANES III and HIS), conditions and exposures relevantto occupationalhealth should be included.This processwill requireclose collaborationbetween NIOSH, other federalagencies,and other groupsthat will use these data in the future. In utilizingexisting health data sources,the advantagesand disadvantages, as discussed above, must be carefully considered. In many instances, existing data which characterizeworkplace exposures, clinical outcome, and vital recordsare extremelyuseful in ongoing surveillance activities.In that context, furtherwork is underway to evaluatethe utility of certain data sources and will be summarizedelsewhere (z8). In performing this work, coordination and integrationof these existing data sourcesinto ongoing programsof preventionof occupationalillness and disease is essential. Too often, data are collected in isolation and not linked directlyto preventionprograms. If surveillanceis to be used to guide preventionactivities,such linkage is essential. CONCLUSION

The purposeof surveillanceis to targetwork-relateddiseasesand injuries so thatpreventioncan occur (z9). Therefore,the utilityof anysurveillance effort must ultimately be judged by one criterion: "Does the program contributeto the preventionof occupationaldiseasesand injuries?"Thus, surveillanceis centralto the mission of NIOSH and its role in exercising national leadershiptowardpreventionof work-relatedillnessand injury. In the future,NIOSH will continue to build upon activitiesdescribedin this paperto improvethe capacityfor surveillanceof occupationalillness and injuryin the United States. We appreciatethe participationin the reviewprocessof Acknowledgments: effort,particularly theNIOSHsurveillance manycolleagueswhohaveevaluated Coordinating Group:ToddM. Frazier, themembersof theNIOSHSurveillance M.D.,PatriciaA.Honchar, Sc.M.,MichaelB.Moll,Ph.D.,ThomasB.Richards,

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

2I9

Ph.D., MurrayL. Cohen,Ph.D., and TheodoreJ. Meinhardt,Ph.D.We also wish to acknowledgeparticularlythe assistanceof Dr. StevenThackerof the wereof Centersfor DiseaseControlwhose commentsand conceptualizations directbenefitto the authors.Othercontributions,too numerousto mention here, will be acknowledgedindividuallyin a monographon Surveillanceof OccupationalDiseaseandInjuryto be publishedin a majorU.S.journal.

REFERENCES Langmuir,A. D. "William Farr:Founder of Modern Concepts of Surveillance," Intl.J. Epid. 5 (I976): I3-I8. z. CDC. "CDC Comprehensive Plan for Epidemiologic Surveillance,"U.S. Department of Health and Human Services,Public Health Service,Centers for Disease Control, I987. 3. Thacker,S. B., Choi, K., and Brachman,P.S. "The Surveillanceof Infectious Disease, "J. Am. Med.Assoc. z49 (i983): ii8i-85. 4. Foege, W H., Hogan, R. C., and Newton, L. H. "SurveillanceProjectsfor Selected Diseases," Intl.J. Epid. 5 (I976): z9-37. 5. CDC. "LeadingWork-RelatedDiseases and Injuries-United States: OccuI.

pationalLungDiseases," M.M.WR. 3z

(i983):

z4-z6,

3z.

6. Harrington,J. M. "EpidemiologicStudy of Work-RelatedDisease," Scand. J. WorkEnviron. Health IO (I984): 3 53-59. 7. Rutstein, D. D., Mullan, R. J., Frazier,T. M., Halperin, W E., Melius, J. M., and Sestito, J. P. "SentinelHealth Events (Occupational):A Basis for PhysicianRecognition and Public Health Surveillance,"Am. J. Pub. Health 73 (i983):

I054-6z.

8. Wegman, D. H., and Froines, J. R. "SurveillanceNeeds for Occupational Health," Am.J. Pub.Health7 5 (i 9 85): Iz59-6i. 9. Froines, J. R., Dellenbaugh, C. A., and Wegman, D. H. "Occupational Health Surveillance:A Means to IdentifyWork-RelatedRisks,"Am. J. Pub. Health76 (I986): I089-96. io. Sundin, D. S., Pedersen, D. H., and Frazier,T. M. "Occupational Hazard and Health Surveillance,"(Editorial),Am.J.Pub.Health76 (I986): I08384. ii. Pederson,D. H., Young, R. O., and Sundin,D. S. "AModel for the Identification of High Risk Occupational GroupsUsing RTECSand NOHS Data," U.S. Department of Health and Human Services, Public Health Service, Centersfor Disease Control, National Institutefor Occupational Safetyand Health,D.H.H.S.,NIOSH, I983, PublicationNo. 83-II 7. iz. CDC. "LeadingWork-RelatedDiseases and Injuries-United States: Severe Occupational TraumaticInjuries,"M.M.WR. 33 (i984): zI3-I5.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

JOURNAL

220

OF PUBLIC

HEALTH

POLICY

* SUMMER

I988

I 3. Silverstein,B. A., Fine,L.J., and Armstrong,T.J. "CarpalTunnelSyndrome:

Causes and A PreventiveStrategy,"Seminarsin Occupational Medicine i (i986): zI3-2I.

Discher,D. P., Kleinmen,G. D., and Foster,F.J. "Pilot Study for Development of an Occupational Disease SurveillanceMethod," U.S. Department of Health, Educationand Welfare,PublicHealth Service,Centerfor Disease Control, National Institute for Occupational Safety and Health, DHEW, NIOSH, I975, PublicationNo. 75-I6z. I 5. Seligman,P.J., Halperin,W E., Mullan, R. J., and Frazier,T. M. "Occupational Lead Poisoningin Ohio: SurveillanceUsing Workers'Compensation Data," Am. J. Pub. Health 76 (I 9 86): I z99-I 30z. i6. Baker,E. L., Landrigan,P. J., Barbour,A. G., Cox, D. H., Folland, D. S., Ligo, R. N., and Throckmorton,J. "OccupationalLead Poisoning in the United States: Clinical and Biochemical Findings Related to Blood Lead I4.

Levels," Brit. J. Ind. Med. 36 (I979):

3 I4-zz.

Muldoon, J. T., Wintermeyer,L. A., Eure,J. A., Fuortes, L., Merchant,J. A., Vanlier,S. F., and Richards,T. B. "State Activities for Surveillanceof OccupationalDisease and Injury,I985," SurveillanceSummaries,U.S. Departmentof Health and Human Services,PublicHealth Service,Centersfor Disease Control, I987. I 8. Halperin,W E., and Frazier,T.M. "Surveillance for the Effectsof Workplace Exposure,"Ann. Rev. Pub. Health 6 (I985): 4I9-3 z. I9. Frazier,T. M. "OccupationalHealth Reporting Systems-U.S.A," Am. J. Ind. Med. 8 (I985): 44I-46. zo. Rosenstock, L. "Hospital-Based Academically Affiliated Occupational I7.

Medicine Clinics," Am. 2I.

J. Ind. Med. 6

(I984):

I55-58.

Blanc, P. D., and Olson, K. R. "OccupationallyRelatedIllnessReportedto a Regional Poison Control Center,"Am. J. Pub. Health 76 (I986): I 30307.

zz. Levy,B. S. "TheTeachingof OccupationalHealth in UnitedStatesMedical Schools: Five-YearFollow-Up of an Initial Survey,"Am. J. Pub. Health 75 (I985): 79-8I.

z3. NIOSH. "SentinelEventNotification Systemfor OccupationalRisks (SENSOR) ... A Proposal,"U.S. Departmentof Health and Human Services, Public Health Service,Centers for Disease Control, National Institutefor OccupationalSafety and Health, I987. z4. Ferris,B. G. "EpidemiologyStandardizationProject,"Am. Rev. Resp. Dis. ii8 (6, part 2) (I978): I-iz0. z5. Medical ResearchCouncil'sCommitteeon the Aetiology of ChronicBronchitis. "StandardizedQuestionnaireson RespiratorySymptoms,"Brit.Med. J. z (I960):

I665-70.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions

BAKER

ET AL. * OCCUPATIONAL

ILLNESS

AND

INJURY

22I

z6. Rose,V E. "Reliability andUtilizationof OccupationalDiseaseData,"U.S. of Health,EducationandWelfare,PublicHealthService,Center Department forDiseaseControl,NationalInstitutefor OccupationalSafetyandHealth, DHEW,NIOSH, 1977, PublicationNo. 77-I 89. 2.7. Thacker, S. B., Parrish,R. G., Trowbridge, F.L., andthe Surveillance CoordinationGroup."AMethodto EvaluateSystemsof Epidemiologic Surveillance,"U.S. Departmentof Health and Human Services,PublicHealth Service, Centers for Disease Control, I987.

z8. Baker,E. L., ed. Surveillance of OccupationalDiseaseand Injury,Monographin Preparation. 29. Millar,J. D. "Screening andMonitoring:Toolsfor Prevention," J. Occup. Med. 2.8 (I986): 544-46. ABSTRACT Surveillanceof occupationalillnessand injuryis essentialfor targetingworkplace preventionefforts. Surveillancesystems should include mechanismsfor standardized data collection, data analysis, and dissemination of results to "all who need to know." In occupational health, surveillancesystems are less developed than in communicable disease prevention.The disparityexists because occupational health surveillanceis conceptually more complex and is limited by legal and social impediments. The National Institute for Occupational Safety and Health (NIOSH) has listed improvementin systems for occupational illness and injury surveillance as a top organizational priority. In this paper, we review current approaches to surveillance designed to identify cases of occupational illness or injury or to monitor secular trends. Gaps in the surveillanceeffort include the absence of a functional system for practitionerreportingof selected occupational disordersand a standardapproach to collection of certain health data. NIOSH programs designed to addressthese limitations include a Sentinel EventNotification Systemfor OccupationalRisks (SENSOR)which links health care providers with state health departmentsfor the purpose of reportingand follow-up of cases of occupational illness and injury.Other new programs are designed to improve the quality of existing data sources for use in surveillance and to develop new approaches to data collection.

This content downloaded from 138.253.100.121 on Sat, 05 Dec 2015 18:13:45 UTC All use subject to JSTOR Terms and Conditions