findingsonthe serologictest for syphilisinadults › nchs › data › series › sr_11 ›...

NATIONAL CENTER Series 11

For HEALTH STATISTICS Number9

VITAL and HEALTH STATISTICSDATA FROM THE NATIONAL HEALTH SURVEY

Findingsonthe

SerologicTest for

Syphilisin Adults

United States. 1960-1962

A discussion of the serologic tests for syphi I is used,

with data on the percent reactive to the Kolmer

Reiter Protein test by age, sex, and race, and an

analysis of differentials by place, income, educa-

tion, occupation, and marital status.

DHEW Publication No. (HRA) 74-1275

U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFAREPublic Health Service

Health Resources AdministrationNational Center for Health Statistics

Rockville, Maryland

Vhl and Health Statistics–Series 11, No. 9Reprinted as DHEW Publication No. (HRA) 741275

August 1973

First issued in the Public Health Service Publication Series No. 1000, June 1965

COOPERATION OF THE BUREAU OF THE CENSUS

In accordance with specifications established by the Na-tional Health Survey, the Bureau of the Census, under a con-tractual agreement, participated in the design and selection ofthe sample, and carried out the first stage of the field interview-ing and certain parts of the statistical processing.

Library o{ Congress Catalog Card Number 6-5-60070

CONTENTSPage

Introduction ----------------------------------------------------------

The Serologic Tests for S~hilis ----------------------------------------

Differences in TestResults ---------------------------------------------

Findings on the KRPTest ----------------------------------------------Age, Race, and Sex--------------------------------------------------Other Demographic Variables ----------------------------------------

Area ------------------------------------------------------------Race and Region -------------------------------------------------Occupation of Employed -------------------------------------------Education -------------------------------------------------------Family Income ---------------------------------------------------Marital Status ---------------------------------------------------

Summary -------------------------------------------------------------

References -----------------------------------------------------------

Detailed Tables -------------------------------------------------------

Appendix I. Serologic Tests for Syphilis ---------------------------------Kolmer Reiter Protein Test ------------------------------------------VDRL Tests --------------------------------------------------------Quality Control -----------------------------------------------------Anticomplementary Specimens ----------------------------------------Other Artifacts -----------------------------------------------------

Appendix II, Statistical Notes ------------------------------------------The Survey Desit~--------------------------------------------------Reliability in Probability Surveys -------------------------------------Missing Data on Examined Persons -----------------------------------Sampling and Measurement Error -------------------------------------Expected Values ----------------------------------------------------Small Numbers -----------------------------------------------------

Appendix 111. Demo~aphic Terms --------------------------------------

1

1

3

557788899

10

10

11

202020202020

23232323252528

29

IN THIS REPORT are presented findi~s on the serologic test for syph-ilis done duving Cycle I of the Health Examination Survey. Cycle I con-sisted of examinations of a nationwide probability sample of persons 18-79yeavs of age, selected from the U.S. civilian, noninstitutional population.

Blood specimens were taken, ad the Kolmev Reiter Protein and Vene-~eal Disease Research Laboratory methods were used to detevmine se-YO1OD”Cevidence of syphilis.

This report describes the serolo~”c tests for syphilis, specifies the tech-niques used, presents the data collected, and compares the informationobtained in this survey with that obtained in othev surveys. The relation-ships of the Kolmer Reiter Protein testfing!ings to the demographic var-iables of age, sex, race, family income, education, place, marital statws,and occupation are examined.

The prevalence of positive STS findings was higher in men than in womenboth in the white and Negro populations. The rates for the Negro popula-tion were substantially gyeater than those for the white population in bothsexes and in evevy age group. The prevalence of reactivity to the sero-logic test for syphilis also vayied by Yegion. The rates weye highest inthe West and lowest in the Northeast fo~ all Yace-sex groups.

SYMBOLS

Data not available ------------------------ --- ICategory not applicable ------------------ . . .

Quantity zero ---------------------------- -

Quantity more than O but less than 0.05 ----- 0.0

Figure does not meet standards ofreliability or precision ------------------ *

.

FINDINGS ON THE SEROLOGIC TESTFOR SYPHILIS IN ADULTS,,

Tavia Gordon and Brian Devine, Division of Health Exarninution Statistics

INTRODUCTION

The National Health Survey uses three meth-ods for obtaining information about the health ofthe U.S. population. The first is a household in-terview in which persons are asked to give infor-mation relating to their health or to the health ofother household members. The second is the col-lection of data from available health records. Thethird is direct examination. The Health Examina-tion Survey (HES) was organized to use the thirdprocedure, drawing samples of the population ofthe United States, and by medical examination andwith various tests and measurements undertakingto characterize the population under study.

The initial enterprise of the Health Examina-tion Survey was the examination of a nationwideprobability sample of 7,710 persons aged 18-79years. Its purp&e was to obtain information onthe prevalence of cardiovascular disease, diabetesand certain other chronic diseases, on dentalhealth, and on the distribution of a number of an-thropometric and sensory characteristics. A bloodspecimen was taken for various tests, includingthree serologic tests for syphilis (STS). Alto-gether, 6,672 persons were examined during thecourse of the Survey, which was begun in October1959 and completed in December 1962. Samplepersons received a standard examination, lastingabout 2 hours, performed by medical and otherstaff members of the Survey in specially designedmobile clinics.

This report deals with findings on the serol-ogic tests for syphilis. It is one of a series of re-ports describing and evaluating the plan, conduct,

and findings of the first cycle of the Health Exami-nation Survey. The descriptions of the generalplanl and of the sample population and response 2have been published. These provide general back-ground for all reports of findings. This reportdescribes the serologic tests for syphilis, spec-ifies the techniques used, presents the data col-lected, and compares the information obtained inthis Survey with that obtained in other surveys.The relationships of test findings to the demogra-phic variables of age, sex; race, family income,education, place, marital status, and occupationare discussed. Comparisons by race are re-stricted to the white and Negro populations, sincethe sample is too small to permit adequate repre-sentation of other nonwhite groups.

THE SEROLOGIC TESTS

FOR SYPHILIS

A venous blood specimen of 15 cc. was takenfrom each examinee. Three cc. were transferredto a blood glucose specimen tube. The remainderwas centrifuged, and 2 to 3 cc. of the serum weretransferred to an STS tube, the remaining serumbeing divided between tubes for serum cholesteroland serum bentonite flocculation determinations.The STS specimens were promptly refrigeratedtwice a week the accumulated specimens wereshipped unrefrigerated to the Venereal DiseaseResearch Laboratory (VDRL) in Chamblee, Geor-gia, for determination of serologic evidence ofsyphilis by the Mazzini, the Kolmer Reiter Pro-tein (KRP), and the VDRL methods. The technical

1

assistance and adviceBranch, Communicableof State Services, U.S.

of the Venereal DiseaseDisease Centerj BureauPublic Health Service is

gratefully acknowledged, the more so because oftheir recognized leadership in the area of serol-ogic tests for syphilis.

The serologic tests were performed and re-ported in accordance with techniques describedin the manual of Serologic tests for Syphilis— 1964(Appendix I). The Mazzini and VDRL tests use themore sensitive cardiolipin antigen, while theKRP test uses the more specific treponemalantigen. This multiple determination allows acomparison of the results of different tests onthe same individual as well as of, the overallresults of the different tests. Only the resultsof the KRP and VDRL tests are presented inthis report.

To use the data in this report, it is necessaryto have some understanding of the essential fea-tures of the serologic test for syphilis and anawareness of the limitations in our understandingof the immunological mechanisms on which thesetests are based. While an exhaustive review ofthis subject is out of the question here, a few re-marks may be helpful.

Syphilis is caused by a treponeme-Trep-onema pallidum. The normal response to infectionby this organism is an increase in the concentra-tion of an antibody in the blood called reagin. Theclassical serologic tests for this infection, begin-ning with the Wassermann and including the VDRL,determine the presence or the concentration ofthis substance. If effective treatment is begunearly enough, it will normally reduce the concen-tration of reagin below a level measurable by thetests in use. Even without treatment the concen-tration of antibody will tend to drop with the pas-sage of time. It is impossible, however, to deter-mine from the reagin concentration whether thesyphilis is early or late, whether it was contractedrecently or a long time ago, or whether it is activeor cured.3

Like all biological phenomena, antibody de-velopment and decline is highly variable from per-son to person, both with respect to time and level.A few infected persons never develop reagin; sometreated persons never lose reagin. Some personsdevelop reagin without having contracted syphilis,and there is a long roster of illnesses which have

been shown to cause transient rises in this anti-body. Moreover, some persons maintain per-manently elevated concentrations of reagin as aconsequence to some of the collagen diseases orfor other reasons. In other words, reagin is nota specific treponemal antibody.

Beginning in 1949 with the Treponema Pal-lidum Immobilization test, a new group of testswas developed which determined the presence orconcentration of a specific treponemal antibody inthe serum. Included among these tests is theKRP test, which uses as antigen a protein frac-tion of the Reiter treponeme. The treponemalantibody appears at about the same time afterinfection by this organism as does reagin, butthe rise and fall of antibody level after infectiondcxas not necessarily follow the curve for reagin.Present evidence suggests ‘that this specific anti-body persists at a high level much longer thanreagin and is less affected by treatment.

The KRP test has the usual limitations. Thereare occasions when the KRP testis positive in theabsence of syphilis. Like all the serologic testsfor syphilis it cannot distinguish between syphilisand other treponemal diseases (such as yaws).What is more, there are certain categories orstages of syphilis in which the K~P test is lesssensitive or less likely to yield a positive re-action than are the reagin tests, such as theVDRL. Thus, the KRP is neither completely spe-cific nor completely sensitive.

Also, it is not diagnostic pev se. In individualsno serologic test for syphilis is in itself diagnosticof the disease. What is necessary in addition tothe serologic test is (1) a careful physical exami-nation with syphilis in mind, (2) a history specif-ically directed to both syphilis and other diseaseswhich may produce positive serologic tests forsyphilis, and (3) an investigation of case contacts.Furthermore, adjunct laboratory procedures arefrequently indicated. The Health Examination Sur-vey did not include any examination of the geni-talia or the perianal region in its physical exami-nation, nor did the medical history inquire aboutsyphilis or related matters. There was also QOcase followup.

Despite the limitations of the HES examinationand of the serologic test for syphilis as indicatorsof the disease, the findings reported here can beregarded as indicating not only the extent to which

persons in this population have been exposed tosyphilis but also the socioeconomic differentialsof this disease. While apositive serologic test forsyphilis may not in itself be used for diagnosingthe disease in an individual, it still points to thestrong probability y of the disease. Moreover, evenin clinical practice the STS must often be thecourt of last resort in diagnosis because of theabsence or ambiguity of other evidence. Finally,in the findings of the prevalence of this diseasethe cases incorrectly diagnosed as syphilis by theserologic test are balanced against the cases ofsyphilis missed. What the net balance will bevaries with the test under consideration, the pop-ulation being tested, and other factors.

DIFFERENCES IN TEST RESULTS

In this report a test result is consideredpositive whether the reactions weak or strong.However, there were sharp differences among thetests with respect to the ratio of reactive to weaklyreactive results (table A). A large majority ofthe positive KRP results were reactive; therewere seven reactive tests to each weakly reactivetest, In the VDRL test the situation was reversed,with a ratio of weakly reactive to reactive of 3 to1. While weakly reactive VDRL tests are morecommon than reactive ones in both the white andNegro populations, the ratio of weakly reactiveto reactive is less in the Negro population thanin the white. For the VDRL tests, then, positivereactions were, on the average, stronger for the

Negro population than for the white; this reflectsa higher average concentration of reagin in theNegro. This may indicate either a higher pro-portion of relatively new infections or a higherproportion of untreated cases. For the KRP testrace differential is evident in the ratio of weaklyreactive cases for women but not for men.

The prevalence rates of reactivity for adultsby sex for each of the serologic tests for syphilisare given in table B. In the KRP test the rate formen is distinctly higher than that for women. Inthe VDRL test the rates are essential y the samefor both sexes. However, if weakly reactive re-sults are omitted, the rate for men is higher thanthe rate for women in the VDRL as well as theKRP test.

Both tests show substantially higher prev-alence rates for the Negro population than thewhite. This is true whether the reactive tests orthe weakly reactive tests are combined or con-sidered separately. However, the preponderanceof the rate for Negroes is distinctly less if thetests are combined than if only reactive VDRLtests are considered. The largest race differentialis manifested by the KRP test.

It seems reasonable to attribute these testdifferentials to differences in specificity, with theKRP being more specific than the VDRL test—thatis, less likely to show a positive reaction in theabsence of syphilis. The explanation, however,may well be more complex.

In general, both tests show the same trendsby age (fig. 1). The chief point of difference is that

Table A. Prevalence of reactive and weaklv reactive serologic tests for syphilis inadults, by race, sex, and test-used: United States, 1960-fj2

KRP VDRL

Race and sex

Reactive Weakly Reactive Weaklyreactive reactive

Number of positive reactions per 100 adults inspecified group

White men---------------------------- 2.0 0.3 0.9White women--------------------------

2.50.5

Negro men0.5

-------- -------- -------- ---- 2A:; 3.0 6.3Negro women -------------------------- 16.1

1?::0.3 5.4 5.5

NOTE: Sample too small to permit adequate representation of other races.

3

. .Table B. Prevalence of reactive and weakly

reactive serologic tests for syphilisin adults, by sex and test used: UnitedStates, 1960-62

I I

Test Bothsexes Men Women

Number of positive reac-~ tions per 100 adults in

specified group

Total---w~

React3-ve---- 3.5 3.8 3.1Weakly reac-

tive ------- 0.5 0.6 0.5

VDRL I II ITotal---

1+~Reacti_ve---- 1.2 1.5 1.0Weakly reac-

tive ------- 3.5 3.3 3.6

for men the prevalence ofpositive serology con-tinuesto rise through theentireagerange with theVDRL test, whereas with the KRP test the risecontinues to ages 45-54 years, after which thereis a decline. The parallelism of these two testsis generally better ifweakly reactive VDRLtestsare omitted from the comparison, but even withthis omission the test differences with respecttothe age trend for menat older ages remain. Theage trends for whiteandNegropersons areessen-tially the same , with positive reactions being sub-stantially higher among the Negro atallages(fig.2). This istrueeven ifweaklyreactive VDRL testsare excluded, although the race differential forthis testis then accentuated at all ages.

From the preceding discussion it is evidentthat the two tests present differing pictures ofthe prevalence of sypbilis and of differentials byage, race, and sex; but these pictures are notnecessarily inconsistent.

These test comparisons maybe concluded bya consideration of the agreements and disagree-ments between the tests. In table C the level ofagreement and disagreement by sex and race isexhibited. In most cases. thelower the percentageof positive findings, the higher the agreement be-fween the two tests. This reflects the large

amount of agreement with reSpect to negativefindings. However, the proportion of positivecases on which the two tests agree is greaterthe highey the prevalence. Thus, the VDRL and

PERCENT

“~VDRL KRP— _ Mm

m— -—. women

10 –

8 –

I

6 —

4 –

o I I I I I I20 so 40 m 60 70 80

AGE IN YEARS

Figure 1. STS prevalence rate by age and sex,both tests.

4

PERCENT

‘00’0~80,0

60.0

E

— Men.--- worm”

40s0 !-

AGE IN YEARS

Figure 2. STS prevalence rate by age, race, am!sex; KRP test.

the KRP are less likely to be in overall agree-merit in the Negro population thanin thewhite butare more likely tobe in agreement with respectto cases considered positive by either test.

FINDINGS ON THE KRP TEST

Throughout the remainder ofthis report thediscussion will be restricted to the findings ofthe KRP test. Weakly reactive and reactive find-ings will be grouped together under thelabel’’re-

active. ” It is believed that, in general, this testprovides minimum estimates of the prevalence ofsyphilis. By syphilis is meant an infection with T.pallidum which has occurred sometime in the lifehistory of the individual. This infection may haveoccurred many years before the test was admin-istered or more recently. It may or may not havebeen cured; it may or may not have producedsecondary effects evident on physical examination.The KRP is chosen in the belief that it is nearlyas likely to be positive as the VDRL test if theperson has had syphilis and less likely to be pos-itive if a person has not had syphilis.

Ager Race, and Sex

To review briefly the ‘findings with respectto the age, race, and sex on the KRP: it isestimated that more than four million adultsare reactive to the KRP test, a rate of 4.0 per-cent (tables 2 and D).

The prevalence of positive findings is higherin men than in women, both in the white and theNegro populations. The rate for men rises to ages35-44 and remains fairly constant through ages65-74, dropping sharply at ages 75-79 years. Therate for women increases to ages 55-64 and thensharply decreases. The peak age-specific ratesfor men and women are about the same. Ratesfor the Negro population are substantially greaterthan those for the white population in both sexesand in every age group (table 1). The trend of prev-alence with age is remarkably similar in the tworaces. The sex differentials in prevalence amongthe white population vary erratically from one agegroup to the next, being higher for women thanmen in some age groups, but, on the average,slightly higher for men. In the Negro population,where a larger sex differential is noted, the prev-alence rate for men is higher than that for womenin every age group except the age group 55-64.

Previous studie~ ,h~ve arrived at largelycomparable “results. ‘ The percentage of re-active specimens increases with age, althoughthe studies differ in the age at which the highestpercentage is reached. Where data are availablefor both sexes, the percentage is higher for menthan for women.. Where data are available forboth white and Negro persons, the race differentialis similar to that found in the Health ExaminationSurvey.

5

Table C. Percent distribution of agreement between KRP and VDRL test results, by raceand sex: United States, 1960-62

All races White NegroKRP VDRL

result resultMen Women Men Women Men Women

Percent distribution of all determinations i.nspecified group

All test results -------- 100.0 100.0 100.0 100.0 100.0 100,0

Agreement --------------- 95.1 94.1 96.4 94.9 83.9 87.2

Positive--Positive-- 13.0Negative--Negative-- 92:: 9;:: 9;:: 9;:: 70.9 7;:;

Disagreement ------------ 4.9 5.9 3.6 5.1 16.1 12.8

Negative--Positive-- 2.7 3.5 2.4 3.4 3.6Positive--Negative- - 2.2 2.4 1.2 1.7 12:: 9.2

NOTE: Sample too small to permit adequate representation of other races.

Table D. Number and percent of adults re-active to the KRP test for syphilis, byrace and sex: United States, 1960-62

RaceBoth Mensexes

Women

Number of adults inthousands

All races--- H!==+=White ------------- 2,136 I 1,047 1,089Negro ------------- 2,206 1,191 1,015

Percent reactive inspecified group

All races---

H*White -------------Negro ------------- 1;:: I 2;:; 1;::

NOTE : Sample too small to permit ade-quate representation of other races.

The interpretation of these differentials is acomplicated matter. It may be assumed that thesex and race differentials mainly reflect differ-ences in rates of infection. However,ifwomen aremore likelyto obtain earlyor effective treatmentthan men, infection is less likely to leave a se-rologic trace, and the sex differential in rate ofinfection willbe exaggeratedbythe serologic test.A similar exaggeration of the racial differentialwould result if white persons were more likelyto obtain early or effective treatment than Ne-groes.

Age differentials reflect an even more com-plicated set of factors. Asa person grows older,the chance that he will have been infected in-creases. Age differences, then, will reflect ac-cumulated experience. However, they will alsoreflect differences in exposure and treatment,which has varied from one generation to another.A 45-year-old person included in this Surveywill have been 25 years old sometime between

1940 and 1942 when the prevalence of syphilis andthe quality of treatment differed greatly fromthat of 1960-62. Other factors, such as the lossof reactivity with time after infection and the in-creasing likelihood of other illness capable ofproducing false positive reactions, are also in-volved. It is impossible to measure the effectof these various factors with any precision, justas it is impossible to state with any precisionwhat proportion of the positive serologic reactionsrepresent cured infections and what proportionrepresent still active infections.

Other Demographic variables

In the discussions that follow, the populationis classified in a variety of ways—by family in-come, by education, etc. —and the prevalence ofpositive KRP findings in different groups iscompared. In making these comparisons, allow-ance must be made for the fact that there aredifferences from one group to another in the dis-tribution of people by age, ‘race, and sex, and thatprevalence of reactivity to serologic tests forsyphilis varies by age, race; and sex. Becausethe sampling variability of age-race-sex-specificvalues for each group is usually very large, asummary comparison was thought preferable tothe presentatio~ of prevalence rates specific byage, race, and sex. For this reason the actualprevalence rate for each group is compared withan expected rate. The expected value is obtainedby weighting age-race-sex-specific rates for thetotal United States by the age-race-sex distribu-tion for that group under consideration. The ob-vious meaning can be attached to differencesbetween actual and expected rates, with the under-standing that small differences may arise bychance. A positive difference, for example, in-dicates that the prevalence rate for the group ishigher than expected. Alternatively the data canbe presented as a ratio of actual to expected rates.If the ratio is greater than 1.0, the actual rate ishigher than expecte~ if the ratio is less than 1.0,the actual rate is less than expected. Where thereis no statistically significant difference betweenthe actual and expected values for a group, dif-ferences for individual age-race-sex groups usu-

ally exhibit only random fluctuations. The follow-ing discussion is generally restricted to the totalpopulation by sex, but race-specific data are in-cluded in the detailed tables.

Area

The findings in tables 3-6 can be briefly sum-marized. The prevalence rate is significantlyhigher in the West than in the remainder of thecountry (table 3, fig. 3). Otherwise, there are nodiscernible place differences, whether contrasts

Men Women I

N:Or~- South wad N:::I- %uth Wwt

Fi sure 3. Ratio of actual to expected percent re-;ctive to the KRP test for syphi1is”in adults,by region.

7

M?ween population-size groups, urban-rural dif-ferences, or differences between standardmetro-politan statistical areas and other urban placesare considered. Such differences undoubtedly doexist, but they are too small to be determinedfrom the HES data.

Vonderlehr and Usilton found lower rates forthose Selective Service inductees from rural areasthan for those from urban areas ,4 but they foundno discernible pattern to rate differences forinductees by city size.5

Race and Region

The large differences in racial distributionby region and the large racial differences in ratesmake it desirable to consider the regional con-trast separately for Negroes and for white per-sons. Table 6 presents the actual and expectedprevalence rates by race and region. The ex-pected rates are race-sex-adjusted expectedrates.

Regional differences, as age and racediffer-ences, are both large and consistent. The prev-alence rate for every race-sex group was bighestin the West and lowest in the Northeast. Indeed,for all age groups the rates for white men andfor white women were higher in the West than inthe Northeast and the South.

HES findings by race and region are largelyin agreement with those reported by Vonderlehrand Usilton4 5 and Karpinos. 6 The y found sub.stantially lower rates for white and Negro Selec-tive Service inductees in the Northeast but showedlittle or no differences between the rates for theSouth and the West. Region was recorded as theplace where the inductee first registered for thedraft and, consequently, does not reflect any laterinterregional migration before the induction ex-amination.

Region, therefore, appears to be a variable inthe prevalence of reactivity to the serologic testfor syphilis.

Occupation of Employed

Comparisons of employed persons by occu-pation show the prevalence rate for men who werefarmers and farm managers to be significantlylower than expected (table 7). Also, the prev-

Fisure 4. Ratio of actual to exDected Dercent re-;ctive to the KRP test for syphi 1is” in women,by occupation of employed persons.

alence rate for women who were clerical andsales workers was significantly lower than ex-pected, while the prevalence for women who wereprivate household. and service workers was sig-nificantly bigher than expected (fig. 4). Whateverother occupational differences exist are eithersmall or occur in categories which exhibit toolarge a variance to deem the difference statis-tically significant.

Bowdoin et al. in a study in Savannah, Geor-gia,’ found in all race-sex-age groups relativelylow prevalence rates for professional, clerical,sales, and kindred workers. Women domes”ticservice workers were the only occupational groupwho showed a consistently high rate.

Education

There are no statistically significant differ-ences between actual and expected prevalencerates by amount of schooling (table 8). This fact

8

does notpreclude theexistence ofdifferencesbyeducation but merely shows any differences whichexist either to be small or to have exhibited toolarge a variance to be considered statisticallysignificant.

These findings differ from those of someother studies. Hill and Mugge8 in a study of Ne-groes in AtIanta and Clark 9 in a study of whiteand Negro persons in Baltimore found that prev-alence rates varied inversely with amount ofschooling, but neither study adjusted for age. Sim-ilarly Bowdoin et al., in their Savannah study’found that prevalence rates moved inversely witheducation for each age-race-sex group.

Familv Income

The actual and expected prevalence rates byfamily income and the differences between themare presented in table 9. The actual prevalencerates by family income are essentially the sameas the expected rates. Whatever income differ-ences exist are either small or occur in cate-

gories exhibiting too great a variance.

Marital Status

The findings in table 10 can be briefly sum-marized. The actual prevalence rates by maritalstatus are essentially the same as the expectedrates. Only for never-married women is there astatistically significant difference between theactual and expected rate, the actual rate beingless than expected (fig. 5). It cannot be said thatno other marital status differences exist, butthose that do exist are eithq small or occur inthose categories which exhibit too great a vari-ance for the difference to be deemed statisticallysignificant. Except in the age group 18-24 years,where never-married persons predominate, thelarge bulk of adults are married.

Figure 5. Ratio of actual to expected percent re-act ive to the KRP test for syph i1is in women, bymari tal status.

Bowdoin et al.,7 Hill and Mugge ,s Clark,g andUsilton et al.lo all discussed marital status differ-ences in prevalence rates, but a different resultwas found in each study. Since both marital statusand prevalence rates varied considerably by ageand since none of these studies gave their resultsin age-adjusted form, their results cannot be val-idly compared with each other or HES rates.

9

1.

2.

3.

4,

5.

6.

7.

SUMMARY

The prevalence of positive STS findingsis higher in men than in women both inthe white and the Negro populations.Rates for the Negro population are sub-stantially greater than those for the whitepopulation in both sexes and in every agegroup.No significant population-size group orurban-rural difference in rates was found.Regional differences are both large andconsistent. The prevalence rates werehighest in the West and lowest in the North-east for all race-sex groups.Various occupational differences were.

found. Men who were farmers and farmmanagers and women who were clericaland sales workers had prevalence ratessignificantly lower than expected, whilewomen who were private household andservice workers had rates significantlyhigher than expected.No significant education or family incomedifferentials were found in HES data.The only marital status difference foundwas for never-married women, who hadsignificantly lower-than-expected rates.

REFERENCESlU s Nauonal Health Suney: Plan and initial program of the. .

Health Examination Survey. Health Statistics. PHS Pub. No. 584-A4. Public Health Service. Washington. U.S. Government Print-ing office, May 1962.

2National Center for Health Statistics: Cycle I of the Health

Examinsrion Survey, ssmple and respnnae, Vital and Health Sta-tistics. PHS Pub. No. 1000-Series 1l-No. L Public Health Serv-ice. Washington. U.S. Grrvemmenr Printing Office, Apr. 1964.

%foore, J. E., and’Mohr, C. F.: Biologically false positive se-rological tests for ayphilia. ]. A.M.A. 150(5~467-473, Oct. 1952.

%onderlehr, R. A., snd Usilron, L. J.: Syphilis sraong men ofdraft age in the United State a. ]. A.M.A. 120(17):1369-1372, Dec.1942.

5Vonderlehr, R. A., end Usilton, L; J.: Syphilis among .eIect-

ees and volunteers. J. A.M.A. 117(16):1350-1351, Oct. 1941.

6Ksrpinos, B. D.: Venereal disease among inductees. Bull. o/

tbs U.S. Ar-my~Med. Dept. 8:806-820, Oct. 1948.7&wdoin, C. D., Henderson, C. A., Davis, W. T., MOCSC)J. W.*

and Remein, Q. R.: .%cioeconomic factors in syphilis prevalence,Savannsh, Georgia. ]. Ven. Dis. in/omr. 30:131-139, May 1949.

8Hi11, M., and Mugge, R.: Syphilis prevalence aa rcl.ted m s’J-

cial factors in a Negro subcnmmunity. Am. ]. Sypb, 38(6} 583,Nov. 1954.

‘Clark, E. G.: Studies on syphilis in Eaatem Health Districtof Baltimore city. w Prevalence in 1939 by race, sex, age, adsocioeconomic status. Am. ]. Sypb. 29(4)455, July 1945.

10Usilton, L. J., Bruyere, P. T., snd Bmyere, M. C.: The fre-quency of positive serologic tests fnr syphilis in relation to occu-pation sndmaritrd status smongmen nf draft sge. J. Ven. Dis. hr/omr.

26:216-222, Oct. 1945.

000

10

DETAILED TABLES

Page

Table 1. Percent reactive to the KRP test for syphilisin adults, by age, race, and sex:United States, l96O-62------------------------------------------------------------12

2. Number of adulta reactive to the KRP test for syphilis,by age and sex:States, l96O-62---------------------------------.---------------------------!~~=~12

3. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, region, and sex: United States, 1960-62-------------------------------------13

4, Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, population-sizegroup, and sex: United States, 1960-62----------------------14

50 Actual and expected percentreactive to the I(RP teat for syphilis in adulta, byrace, place description,and sex: United States, 1960-62-------:------------------15

6. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace~ Place> and sex: United States, 1960-62-------------------------------------- 15

7. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, occupation,and sex: United States, 1960-62---------------------------------16

8. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, education,and sex: United Statea, 1960-62----------------------------------L7

9. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, family income,and sex: United States, 1960-62------------------------------18

10. Actual and expected percentreactive to the KRP test for syphilis in adults, byrace, marital status,and sex: United States, 1960-62-----------------------------19

11

Table 1. Percent reactive to the I(RPtest for syphilis in adults, by age, race, and sex: lJnitedStates, 1960-62

Age

Total-18-79year.s------------

18-24

25-34

35-44

45-54

55-64

65-74

75-79-

years------------------------

yeara------------------------

years------------------------

years------------------------

years-------------------------

years------------------------

years------------------------

II I

-=241 ‘i: NegroI 1

Men I w-n II Men I women I Men I women

4.4

0.5

1.8

6.3

5.7

6.1

6.1

1,6

Percent reactive in specified group

3.6

0.3

2.2

3.6

6.1

6.8

3.2

2.3

0.3

1.0

3.3

2.1

3.5

4.0

1,0

2.1

0.4

1.5

2.0

2.8

4.1

2.4

NO17E: Sample too small to permit adequaterepresentationof other races.

22.9

2.8

7.3

30.1

33.4

31.0

32.6

10.1

16.3

7.0

16.7

31.5

35.2

13.1

Table 2, Number of adults reactive to the I(RPtpst for syphilis, by age and sex: United States,1960-62

Age Both sexes Men Women

Number of adults in thousands

Total-18-79 years------------------------ 4,414 ‘ 2,305 2,109

18-24 yeara--------------- -----------------Q-- 68 39 29

25-34 years------------------------------------ 436 188 248

35-44 years------------------------------------ 1,162 718 444

45-54 years------------------------------------ 1>215 574 641

55-64 years--------.--------------------------- 1,012 462 550

65-74 yeara.........--------------------------- 499 303 L96

75-79 years------------------------------------ 23 23

12

Table 3. Actual and expected percentreactive to the KRP test for syphilisin adults, by race,region, and eex: United States,’1960-62

Race and region

All races

Northeast--------------

south---------”----””--

weft-------------------

White

Northeast--------------

south------------------

West-------------------

Negro

Northeast--------------

South------------------

Weat-------------------

Men Women

Actual Expected Difference Actua1 Expected Difference

Percent reactive in specifiedgroup

2.7

5,8

5.0

1.8

1,9

3.0

14.9

21.4

34,3

3.5

6,4

3.8

2,3

2.3

2.2

21.6

22.6

24.9

-0.8

-0.6

1.2

-0.5

-0.4

0.8

-6.7

-1.2

9.4

2.1

4,5

4.5

1.1

1.5

3.7

14.0

17,0

17.0

3.2

4.9

2.9

2.1

2.1

2.2

16.3

16.6

15.2

-1.1

-0.4

1.6

-1.0

-0.6

1.5

-2.3

0.4

1.8

NOTE: Sample too small to permit adequaterepresentationof other races. For many categories,estimatesfor the Negro populationhave a very high sampling variability. Comparisonsfor thisgroup should be consideredas indicativeonly.

13

Table 4. Actual and expected percentreactive to the KRP test for syphilisin adults, by race,population-sizegroup, and sex: United States, 1960-62

Race and population-size group

All races

Giant metropolitanareas-----------------

Other very largemetropolitanareas----

Other standardmetro-politan statisticalareas-----------------

Other urban areas------

Rural areaa------------

White

Giant metropolitanareas-----------------




Rural areae------------

M23122

Giant metropolitanareaa-----------------




Rural areas------------

Men Women

Actual Expected Difference Actua1 Expected Difference

3.9

2.8

4.7

3.7

6.3

2.9

2.4

2.2

1.9

1.6

12.7

7.6

40.5

16,1

33.0

Percentreactive in specifiedgroup

4.0

3.7

3.6

4.9

5.6

2;3

2.3

2.2

2.2

2.3

20.8

20.5

22.4

23.5

25.2

-0.1

-0.9

1.1

-1,2

0.7

0.6

0.1

-0.3

-0.7

-8.1

-12.9

18.1

-7.4

7.8

3.5

2.4

3.1

3.6

5.5

2.2

1.5

1.’7

1.9

3.5

15.0

11.2

21.5

15.3

18.2

3.7

3.5

3.1

3.7

4.1

2,2

2.1

2.1

2.1

2.1

16.4

16.7

16.0

15,3

17.0

-0.2

-1,1

-0.1

1.4

-0.6

-0.4

-0,2

1.4

-1.4

-5.5

5.5

.

1.2

NOTE: Sample too small to permit adequaterepresentationof other races. For many categories,estimatesfor the Negro populationhave a very high sampling variability. Comparison for thisgroup should be consideredaa indicativeonly.

14

Table 5. Actual and expected percent reactive to the KRP teat for syphilis in adults, by race,place description, and sex: United States, 1960-62

Race and placedescription

All racea

SMSA-in central city---SMSA-outsidecentral city----------

Urban, not SMSA--------Rural, farm------------Rural, nonfarm---------

White

SMSA-in central city---SMSA-outsidecentral citv----------

Urban,Rural,Rural,

not SlkA--------farm------------n~nfarm---------

MEWSMSA-in central city---SMSA-outsidecentral city----------

Urban, not SMSA--------Rural, farm------------Rural. nonfarm---------

Men I Women

Actual Expected Difference Actual Expected Difference

5,8

2.63.0

:::

3.1

2.01,5

;::

20.6

2L.717.818.132.4


5.2

2:5.55.7

2.2

n2.62,2

21.6

20.322.823,526,2

0,6

-0.3-1.0-1.71.1

0,9

-0.3-0.7-1.30.1

-1.0

-5.46.2

4,4

2.2

i:;4.1

2,1

;::4,72.5

16.8

16.614.813.417.6

4.5

2.6

:::3.7

2.1

2.1

;::2.0

16.4

17.714.613,818.2

-0,1

-0.4-0.1

:::

0.4

-1.1

-:::-0.6

SNSA = Standard metropolitan statistical area.

NOTE: Sample too small to permit adequate representation of other races. For many categories,estimates for the Negro population have a very high sampling variability. Comparisons for thisgroup should be considered as indicative only.

Table 6. Actual and expected percent reactive to the KRP teat for syphilis in adults, by race,place, and sex: United States, 1960-62

Race and place

All races

Urban------------------Rural------------------

White

Urban------------------Rural------------------

N!2SxQ

Urban------------------Rural------------------

Men Women



4.34.5

2.41,9

20.826.9

4,24.6

2.22,3

21.625.1

0.1-0.1

-:::

-0.81.8

3.73.4

2.12.2

16.715.2

3.63.5

16.016.7

-!):;

O,i

-’HNOTE: Sample too small to permit adequaterepresentationof other races. For many categories,

estimatesfor the Negro populationhave a very high sampling variability. Comparisons for thisgroup should be considered as indicative only.

15

Table 7. Actual and expected percent reactive to the KRP test for syphilis in adults, by race,

Race and occupation

All races

Professional,technical,and managerial --------

Farmers and farmmanagers --------------

C~~~ and sales---------------

Craftsmen, foremen, andkindred workers-------

O~o;;;Is and kindred----------------

Private household andservice workers-------

Farm and other laborers(except mine)---------

White

Professional,technical,and managerial --------

Farmers and farmmanagers--------------

C~o;~;~ and sales---------------


O~;pe~es and kindred---------------

Private household andservice workers -------


&EzS!

Professional,technical,and managerial--------

Farmers and farmmanagers--------------

C~o;;~~ and sales------...-----.


O+x;;~~s and kindred---------------

Private household andservice workers-------


occupation, and sex: United States, 1960-62

MenI Women



2.6

2.4

3,2

3.1

4.4

3.5

13,6

2.1

1.5

3.1

1.7

2.1

2.1

1.4

23,5

11,7

6.8

23.3

19.8

10.7

36.2

2.8

5.2

2.5

3.4

4.4

5.7

10.4

2.3

2.8

2.2

2.2

1.9

2.2

1.8

27.2

29.0

9.3

22.6

20.8

22.6

25.7

-0.2

-2.8

0,7

-0.3

-2.2

3,2

-0.2

-1,3

0.9

-0.5

0.2

-0,1

-0.4

-3.7

-17.3

-2.5

0.7

-1.0

-11.9

10.5

2.3

*

1.0

*

2.4

.11.0

*

2.1

*

1.0

*

1.4

4.8

*

5.3

*

-

*

12. 5“

23.4

*

3.1

*

2.3

*

3.2

8.0

*

2.4

*

2.0

*

2.3

2,2

*

13,3

*

12.7

*

11.8

19.5

*

-0,8

*

-1.3

*

-0.8

3.0

*

-0,3

*

-1.0

*

-0.9

2.6

*

-8.0

*

-12.7

*

0,7

3.9

*


16

Table 8. Actual and expectedpercent reactive to the KRP test for syphilisin adults, by race,education,and sex: United States, 1960-62

Race and education

All races

Under 5 years--------

5-8 yeers--------------

9-12 years-------------

13+ years--------------

White


5-8 years--------------

9-12 years-------------

13+ yeara--------------

wi?X


5-8 years--------------

9-12 years-------------

13+ years--------------

Men I Women

Actual Expected

7.8

6.0

3,8

1.7

3.7

2.4

2.3

1.4

19.0

29.4

18.8

15.7

IMfferencel Actual I Expected I Difference

Percent reactive in specifiedgroup

10.:

5.7

3.4

2.5

2.8

2.7

2,0

2.1

28.1

25.3

18.0

19,4

-2.:

0’.-

Q.L

-0.8

0.9

+.:

0,3

-0,7

-9.1

4.1

0.8

-3.7

7.8

6.1

2.3

2.5

1.8

2.5

1,9

2.3

26.6

26”.8

6.2

6.4

7.2

5.0

2.8

2.7

2.4

2.5

2.0

2.0

21.9

19.9

11.3

14.4

0.6

1.1

-0.5

-0.2

-0.6

-0,1

0.3

4.7

6.9

-5.1

-8.0

NOTE: Sample too small to permit adequaterepresentationof other races. For many categories,estimatesfor the Negro populationhave a very high sampling variability. Comparisonsfor thisgroup should be consideredas indicativeonly.

17

Table 9. Actual and expected percent reactive to the KRP test for syphilis in adults, by race,family income, and sex: United States, 1960-62

Race and family income

All races

Under $2,000-----------

$2,000-$3,999----------

$4,000-$6,999----------

$7,000-$9,999-””--””””-

$10,000+---------------

Unknown------------:---

White

Under $2,000-----------

$2,000-$3,999----------

$4,000-$6,999----------

$7,000-$9,999----------

$10,000+---------------

Unknown----------------

NE13z!?

Under $2,000-----------

$2,000-$3,999----------

$4,000-$6,999----------

$,7,000-$9,999----------

$lo,ooot---------------

Unknown----------------

Men Women



6.2

7.6

3.7

3.2

1.5

4.7

2.5

1.5

2.5

2,6

1.3

3.1

18.4

34.2

20.1

L7.3

14,8

7.4

6.2

3.4

2,8

2.9

4.8

2.5

2.2

2.1

2.2

2.5

2.3

24.0

23.4

21.5

20.2

21.1

22.9

-1.2

1.4

0.3

0.4

-1.4

-0.1

-0.7

0.4

0.4

-1.2

0.8

-5.6

10.8

-1.4

-2.9

-21.1

-8.1

7.1

3.6

3.0

2.1

2.7

2.9

1.7

2.3

2.1

2.2

2.8

1.3

22,1

12.3

15.5

16.1

6,3

3.9

2.8

2.6

2.2

3,9

2.1

2.2

2.0

2.1

2.2

2.3

18.3

15.0

14,8

13.4

3.0

17.2

0.8

-0.3

0.2

-0.5

0,5

-1.0

-0.4

0,1

0.1

0,1

0,6

-1.0

3.8

-2,7

0.7

-13.4

-3.0

-1.1


18

Table 10. Actual and expected percentreactive to the KRP testmarital status, and sex: United States,

Race and maritalstatus

All races

Married----------------

Widowed---------------=

Divorced---------------

Separated--------------

Never msrried----------

White

Married----------------

Widowed---------------.

Divorced---------------

Separated--------------

Never married----------

&2i?LQ

Married----------------

Widowed----------------

Divorced---------------

Separated--------------

Never msrried----------

f~os~;hilis in adults, by race,

Men Women


Percentreactive in specifiedgroup

3.9

6.5

11.5

14.7

3.7

2.3

3.1

4.4

*

1.5

22.1

27.0

39.9

*

“17,5

4.3

6.3

8.2

16.9

2.3

2.4

3.2

2.6

*

1.2

24.7

25.9

30,3

*

9,4

-0.4

0.2

3.3

-2.2

1.4

-0.1

-0.1

1.8

*

0.3

-2.6

1.1

9.6

*

8.1

3.5

5.1

5.1

8.5

1.0

2.4

2.2

1.9

*

0.1

15.7

25.6

17.5

*

7.0

3.4

5.0

5.6

6.8

1.9

2.1

2.6

2.5

*

1.3

16.6

22,7

20.5

*

6.6

0.1

0.1

-0.5

1.7

-0.9

0.3

-0.4

-0.6

*

-1.2

-0.9

2.9

-3.0

*

0.4

NOTE: Sample too small to permit adequaterepresentationof other races. For many categories,estimatesfor the Negro populationhave a very high sampling variability. Comparisonsfor thisgroup should be consideredas indicativeonly,

19

APPENDIX I

SEROLOGIC TESTS FOR SYPHILIS

The serologic tests for syphilis used irrthis Surveyare briefly described in this appendix. Detailed specifi-cations of these tests are provided in Serolo@”c Tests forSyphilis, 1964, amanual prepared atthe Venereal Dis-ease Research Laboratory, U.S. Public Health Service(PHS Pub. No. 411). Specific page references are includedin this summary.

Kolmer Reiter Protein Test

The test used Reiter protein antigen (p. 45) and em-ployed a Kolmer one-fifth volume qualitative test withserum (pp. 49-52). Test results were reported as non-reactive, weakly reactive, reactive, or anticomple-mentary. Reactive tests were reported as 1+, 2+, 3+tior 4+.

VDRL Tests

The test was a VDRL slide qualitative test withserum (p. 95). If the qualitative test yielded reactive orweakly reactive results, a VDRL slide quantitative testwith serum was undertaken (pp. 96-98), and the greatestdilution at which a reaction was noted was determined.Reactive results were reported in the same reamer asfor the KRP test.

.—

Quality Control

Quality control procedures used were the standardprocedures described in the manual.

Some internal evidence of quality control maybe ob-tained from an examination of the number of reactivespecimens at each of the 42 places at which the surveywas conducted. It is possible that individual places, bychance, might have more or fewer reactive specimensthan average, but a series of such stands would suggestsome change in laboratory standards. No such evidenceis indicated by the data in table I.

Anticomplementary Specimens

Some 188 or 3 percent of the KRP tests were anti-complementary (AC). Computationally these test re-sults were treated in all tables except A and B as if theywere negative. The estimated rate for all adults on thisbasis was 3.97 percent. In tables A and B these speci-mens were treated as undetermined, and rates werecomputed only for determined specimens. ‘l’he rate for

20

all adults on this basis was 4.00. The difference betweenthese two estimates is lost in rounding to one decimal,but for groups with high prevalence the difference doesremain visible.

It is by no means clear what the correct approachshould be. Since VDRL tests were performed on all anti-complementary specimens ,and since VDRL test resultsare highly correlated with KRP test results, it would havebeen reasonable to have used the VDRL test informationto estimate the unknown KRP test results. Unfortunatelythe available data raise serious doubts that this would bea safe procedure. The following VDRL test results wereobtained on the specimens having anticomplementaryKRP tests.

WeaklyTotal reactive Reactive Norma1

Total- 188 11 5 172

White men- 58 1 57White

women---- 112 10 3 99Nonwhite

men------ 10 1 9Nonwhite

wo~n---- 8 1 7

Women were more likely to have anticomplementarytest reactions than men. Women having anticomple-mentary test reactions were more likely to have reactiveVDRL tests than women not having anticomplementaryreactions. While this may reflect a higher rate of syphi-litic infection in the group, it may also reflect a commonartifact producing both anticomplementation on the KRPtest and a positive reaction on the VDRL test. Thispossibility is emphasized by the fact that the number ofanticomplementary reactions among pregnant womenwas four times the average for that age-sex group.

The computational treatment of tie anticomple-mentary results, however, ahnost certainly leads to anunderstatement of the true percentage reactive to theKRP test in the population.

Other Artifacts

The possibility that a reactive STS may be caused byfactors other than syphilis is of some concern whenevaluating the data presented in this report. There is,unfortunately, little internal evidence on this subject, andno external evidence allowing an imputation to the

Table 1. Number of examined adults and number of adults reactive to the KRP test for syphilis, bystand number and color: Health Examination Survey, 1960-62

Stand number

1.------------------------------

2------------------------------

3------------------------------

4 --------- ------------------ ---

5 . . . . . . . . . . . . . . . . -------- ------

6 . . . ----- -------- . . . . . . . . ------

7-..- . . . . -------- -------- ------

8 -------- -------- -------- ------

9 . - - - - - - - - - - - “ - - - - - - - - - -- - ---- .

10------------------------------

11 --------- --------- --------- ---

12------------------------------

13------------------------------

14------------------------------

15 -------- -------- --------- ------

16------------------------------

17------------------------------

18 -------- -------- -------- ------

19 -------- -------- . ..--... ------

20-22 --------- --------- --.e -----

23 -------- -------- -------- ------

24------------------------------

25------------------------------

26------------------------------

27------------------------------

28------------------------------

29------------------------------

30------------------------------

31------------------------------

32------------------------------

33------------------------------

34------------------------------

35------------------------------

36 . . . . . . . . . . . . . . . . . . . . . . . . ------

37--------A ---------------------

38------------------------------

39------------------------------

40 -- - . . -- - .- - . -- - . - -- - - - ------ - -

41------------------------------

42------------------------------

Total sample population

White

128

120

150

155

159

161

148

156

167

166

146

141

153

112

130

142

142

141

154

390

121

128

100

122

103

69

145

15$

134

164

161

159

155

158

124

132

68

155

108

94

Nonwhite

17

321$

6

8

2

16

17

3

15

9

2

47

21

8

1

5

56

32

24

55

52

71

98

10

11

30

4

8

1

7

4

37

23

92

8

43

59

Sample persons reactive to I(RP

White

1

7

1

2

13

8

2

3

3

6

7

6

2

2

1

2

1

4

9

1

3

1

3

1

3

2

7

3

2

3

1

3

3

1

2

1

Nonwhite

1

4

5

2

1

2

5

8

3

1

1

3-

7

3

10

7

12

12

14

1

3

7

2-

1

8

2

5

10

10

21

general population surveyed by the Health ExaminationSurvey.

Two possible factors were considered. The first waspregnancy. The actual number of reactive tests amongpregnant women and the number expected in women ofthe same age and race was

Actusl Expected

KRP------------ 2,0VDRL----------- : 3.4

Another factor considered was the presence of col-lagen disease. Among persons with a pcsitive serum ben-tonite flocculation test for the rheumatoid factor thenumber of reactive tests was

Actua 1 Expected

KRP------------ 12 12.3VDRL----------- 10 15.0

Thus, in neither group were there more reactivetests, either KRP or VDRL, than expected.

—ooo —

22

APPENDIX II

STATISTICAL NOTES

lhe Survey Design

The Health Examination SurVey is designed as ahighly stratified multistage sampling of the civilian, non-institutional population, aged 18-79 years, of the con-terminous United States. The first stage of the plan isa sample of the 42 primary sampling units (PSU’S) from1,900 geographic units into which the United States hasbeen divided, A PSU is a county, two or three con-tiguous counties, or a standard metropolitan statisticalarea. Later stages result in the random selection ofclusters of about four persons from a small neighbor-hood within the PSU. The total sample included 7,710persons in the 42 PSU’S in 29 different States. The de-tailed structure of the design and the conduct of the Sur-vey have been described in previous reports.1 2

Reliability in Probability Surveys

The methodological strength of the Survey derivesespecially from its use of scientific probability samplingtechniques and of highly standardized and closely con-trolled measurement processes. This does not imply thatstatistics from the Survey are exact or without error.Data presented are imperfect for three important rea-sons: (1) results are subject to sampling error, (2) theactual conduct of a survey never agrees perfectly withthe design, and (3) the measurement process itself isinexact even when standardized and controlled. Thefaithfulness with which the study design was carried outhas been analyzed in a previous report.z

Of the total of 7,710 sample persons 86 percent or6,672 were examined. Analysis indicates that the exam-ined persons are a highly representative sample of theadult civilian, noninstitutional population of the UnitedStates. Imputation for the nonrespondents was accom-plished by attributing to nonexamined persons the char-acteristics of comparable examined persons. The spe-cific procedure used 2consisted of inflating the samplingweight for each examined person to compensate for non-

examined sample persons at the same stand and of thesame age-sex group.

The presumption that positive serology was as com-mon in nonexamined persons as examined is by no meansdemonstrated. It is quite possible that a person knowingor suspecting that he had syphilis would be less likely toappear for examination than other sample persons. Thiscan hardly constitute a predominant reason for nonre-sponse, however; and what is more to the point, the prev-alence of positive serology would have to be very highindeed for the prevalence in the total sample to begreatly in excess of the prevalence found in the examinedgroup. For the true prevalence to be one percentage pointgreater than that found in the examined group, one in tenwhite persons not examined and one in three Negroeswould have had to have positive serology—a rather un-likely situation.

Missing Data on Examined Persons

In addition to persons not examined, there weresome persons whose examinations were incomplete inone particular or another. Age and sex were known forevery examined person, but for 173 people either a bloodspecimen was not available or no STS determinationwas made. The extent of missing information is indicatedin table H. The number of determinations by race andsex is shown in table 111.The problem of anticomplemen-tary KRP results is discussed in Appendix I.

In computing rates for any group, persons for whomno STS determinations were available were consideredto have the same age-sex-race specific rates as persons

on whom determinations were made. Since most of themissing data loss resulted from accidental breakage ofSTS specimen tubes or from insufficient blood beingdrawn, it can be assumed that these were random exclu-sions and, therefore, their rates should not differ signif-icantly from ‘the determined rates. Because of the smallnumber of nondetermined specimens the effect on theoverall rates would be slight even if they differed sub-stantially.

23

Table 11, Number of examinedpersons and number for whom no KRP or VDRLby sex and age: Health ExaminationSurvey, 1960-I

Sex and age

Total-----------------------------------------------------

Men—

18-79

18-2425-3435-4445-5455-64;;-;;

18-79

18-2425-3435-4445-5455-64;;-;;

years........-..................--------------------..--.--

years-------,.---------------------------------------.-----yeara-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------

Women

years-----------------------------------------------------

years-----------------------------------------------------years-----------------------------------------------------yeara-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------years-----------------------------------------------------vears---------.-------.------.-----------------------------

determination was made,62

rotal examined

6,672

3,091

41167570354741826572

3,581

534746784705443299

70

No KRP or VDRLdetermination

173

58

11111;

4

Table III. Number of personswith KRP.or VDRL determinationsand number reactive, by color andsex: Health ExaminationSurvey, 1960-62

I

Number of STS determinations

Color and sex

Total KRP reactive VDRL reactive

All persons........-................----------- 6,499 270 300

White men------------------------------------------- 2,621White women----------------------------------------- 2,959 :: 1::Nonwhitemen---------------------------------------- 412 75Nonwhitewomen--------------------------------------

63507 75 52

24

Sampling and Measurement Error

In this report and its appendices several referenceshave been made to efforts to evaluate both bias and vari-ability of the measurement techniques. The probabilitydesign of the Survey makes possible the calculation ofsampling errors. Traditionally the role of the samplingerror has been the determination of how imprecise thesurvey results may be because they come from a samplerather than from measurement of all elements in theuniverse.

The task of presenting sampling errors for a studyof the type of the Health Examination Survey is compli-cated by at least three factors: (1) Measurement errorand “pure” sampling error are confounded in the data;it is not easy to find a procedure which will either com-pletely include both or treat one or the other separately.(2) Ile survey design and estimation procedure are com-plex and accordingly require computationally involvedtechniques for calculation of variancea. (3) Thousands ofstatistics come from the survey, many for subclasses ofthe ~pulation for which there are small numbers ofsample cases. Estimates of sampling error are obtainedfrom the sample data and are themselves subject to sam-pling error, which may be large when the number ofcases in a cell is small or even occasionally when thenumber of cases is substantial.

In the present report estimates of approximate sam-pling variability for selected statistics are presented intables IV-VII. These estimates have been prepared bya replication technique which yields overall variabilitythrough observatio~ of variability among random sub-samplds of the total sample. The method reflects both“pure” sampling variance and a part of measurementvariance,

In accordance with usual practice the interval esti-mate for any statistic may be considered to be the rangewithin one standard error of the tabulated statistic with[18percent confidence or the range within two standarderrors of the tabulated statistic with 95 percent confi-dence.

Expected Values

In tables 3-10 the actual rates of serologic test re-activity for the various demographic variables are com-pared with the exTected. The computation of expectedrates was done as follows:

Suppose. that in an area (say the Northeast) theHt?ulth Examination Survey estimates that there are N+

th.

persons in the i age-sex-race group (i= 1,2. . . . ., 42;sum of Ni = N).

Suppose the Health Examination Survey estimatesthat the rate of serologic test reactivity for the United

State~ in the ith

age-aex-race group is Xi.

Then @e expected rate of serologic test reactivityfor the area is

~2 NiXiNi

Comparison of an actual value for, say, a regionwith the expected value for that region is undertaken onthe assumption that a meaningful statement can be madewhith holds, in some average way, for all persons in theregion. This may or may not be true. The specified re-gion may have higher values for young persons and lowervalues for old persons than are foun”din other regions.In that case an average comparison will obliterate oneor both of these differentials. A similar remark may bemade with respect to values computed for all races to-gether, since relationships found in one race may not befound in another. Some instances will be noted in the de-tailed. tables where the white and Negro differentials arenot the same. In arriving at the general conclusions ex-pressed in the text, an effort was made to consider all thespecific data, including data not included in this repart;but it must be recognized that balancing such evidence isa qualitative rather than quantitative exercise.

The standard error of the difference between anactual and expected value may be approximated by thestandard error of the actual value (table VII). That thisis a reasonable approximation is indicated by the follow-ing data for percent reactive to the KRP test by incomefor white men and women.

Standard error ofdifference

Replicationestimates

With WithApproxi- varying fixed

W Race-sex Q@QIL weights weights

Under $2,000------- White menWhite women

$2,000-$3,999------ White menWhite women

$4,000-$6>999------ ~~ :e”

$7,000-$9,999------ White menWhite women

$10,BOO+----------- White menWhite women

Unknown------------ White menWhite women

;::0.70.60.50.60.7b. 70.71.0:::

::;::?0.60.50.80.51.01.11.61.3

1.20.3

::;0.30.30.40.50.50.9

ii::

Part of the variance of the difference between actualand expected values arises from the use of estimates ofthe population in computing expected values. If the popu-lation values were known or if the estimates weretreated as constants, the v&iance of the difference wouldbe less as can be judged from the column for “fixedweights. ”

The degree of correspondence +tween the approxi-mations and directly computed variances for differencesbetween actual and expected values will vary somewhatwith the variable under consideration, While direct com-putation is very time-consuming, a few other instanceshave been examined.

25

Table IV. Standarderrors in perc~n~ ;~~tiveti the KRP test for syphilisin adults,byage,race,United States, 1960-62

Age

Total-18-79years------------

18-24

25-34

35-44

45-54

55-64

65-74

75-79

years------------------’------

years------------------------

years------------------------

years------------------------

years,-------------.------..--

years------------------------

years------------------------

Total II White I Negro

Men Women Men Women Men Women

Percent

0.3

0.3

0.5

1,3

1.1

1,2

1.5

*

0.3

0.2

0.7

0.7

1.2

1.4

0,8

0.2

0.2

0.5

0.7

0.4

0.7

1.0

*

0.2

0.3

0.4

0,4

0.6

0.8

0.6

2.7

006

2.9

4.5

5.0

4.7

6.5

*

1.5

2.8

2.5

4.7

5.3

3,9

Table V. Standarderrors in number of adults reactive to the KRP test for syphilis, by age and‘sex: United States, 1960-62

Age

Total-18-79years------------------------------------

18-24 years------------------------------------------------

25-34 years------------------------------------------------

35-44 years------------------------------------------------

45-54 years------------------------------------------------

55-64 years------------------------------------------------

65-74 years------------------------------------------------

75-79 years------------------------------------------------—

Both sexes Men Women

Number in thousands

309

27

109

174

182

152

100

*

184

20

56

144

115

92

76

*

169

L7

74

89

128

110

59

.

Table VI. Standarderrors in number of adults reactive to the KRP test for syphilis,by race andsex: United States, 1960-62

Race Both sexes Men Women

All races--------------------------------------------I Number in thousands

!===-% 184 I 169I

~ite ------------------------------------------------------Negro----------------........------------------------------ !!3Z;Other nonwhite--------------------------------------------- *

1202% 162* *

26

Table VII. Standarderrors in percentreactive to the KRP test for syphilisin adults, by speci-fied characteristics,race, and sex: United States, 1960-62

I

Characteristic

Region

Northeast------------------------------------------------South----------------------------------------------------WesC-----------------------------------------------------

Population-sizegroup

Giant metropolitanareas---------------------------------Other very large metropolitanareas----------------------Other standardmetropolitanstatisticalareas------------Other urban areas----------------------------------------Rural areas----------------------------------------------

Place description

SMSA-in central city-------------------------------------SMSA-outsidecentral city--------------------------------Urban, not SMSA------------------------------------------Rural, farm----------------------------------------------Rural, nonfarm-------------------------------------------

Occupationof employed

Professional,technical,and managerial------------------Farmers and farm managera--------------------------------Clericaland salea workers-------------------------------Craftsmen,foremen,and kindred workers------------------Operativeaand kindred workers---------------------------Privatehouseholdserviceworkers------------------------Farm and other laborers (exceptmine)--------------------

Education

Under 5 years--------------------------------------------5-8 years------------------------------------------------9-12 yeara-----------------------------------------------13+ yeara------------------------------------------------

Family income

Under $2,000---------------------------------------------

II

2,000- 3,999--------------------------------------------4,000- 6,999--------------------------------------------7,000- 9,999--------------------------------------------

$lo, oow-------------------------------------------------Unknom --------------------------------------------------

Marital status

Married--------------------------------------------------Widowed--------------------------------------------------Divorced-------------------------------------------------Separated------------------------------------------------Never married--------------------------------------------

White ‘1 Negro

Men I Women I MenIWomen

0.3

:::

0.70,6

::;0.6

::;

:::0.5

0,4

::?

H

::;

2.0

::;0.4

6:;0.50.70.71.4

0.23.13.3

0.;

Percent

0.4

:::

0.5

::;0.61.1

0.7

%;1.40.6

1,0

0.3

;::

H

1.3

:::0.8

0.6

:::0.7

;::

0.40.7

H0.2

2:5.1

5.;

:::4.e11.5

::;6.2

1:::

14.1

:::8.24.0

1:::

6.7

::212.5

1;::

1::!

7.4

4.410.820.012.910.5

3.0

;::3.13.6

1;:;5.94.73.5

4.8*

6.;3.57.9

:::0.94.5

2.47.77.0

%;

27

Specificcharacteristic

Matiitalstatus

Married-------------------Widowed-------------------Divorced------------------Separated-----------------Never married-------------

Marital status

Married-------------------Widowed-------------------Divorced------------------Separated-----------------Never married-------------

Population-sizegroup

Giant metropolitanarea---Other very large metro-politan area-------------Other S14SA----------------Othe~ urban area----------Rural---------------------

Income

Under $2,000--------------

‘1

2,000- 3,999-------------4,000- 6,999-------------

!;60:;()+9,.999-------------

------------------Unkkwm-------------------

Race

WhiteWhiteWhiteWhiteWhite

NegroNegroNegroNegroNegro

White

WhiteWhiteWhiteWhite

WhiteWhiteWhiteWhiteWhiteWhite

Men

Approximation

0.23.13.3

0:7

12;20.012.910.5

0.7

0.60.90.70.6

::;0.50.70.71.4

Replicationwtf.matewith

varyingweights

0.2

H’0.60.8

d:;11.711.27.3

0.7

0.70.90.70.6

;::0.60.81.01.4

Women

approximation

0.40.7;.;

0:2

::;7.08.23.5

0.5

0.60.90.61.1

0.60.60.60.7

u

Replicationestimatewith

varyingweights

0.20.71.2

::!

0.5

0.60.80.61.1

0.90.5

:::1.11.3

Small Numbers Obviouslyinsuchinstancesthestatistichasno meaninginitselfexcepttoindicatethatthetrueauantitvissmall.

ln some tablesmamitudesareshownforcellsfor Suchnumbers,ifshown,havebeeninciudedt;conveyanwhichsample sizeiss;smallthatthesamplingerror impressionoftheoverallstoryofthetable. -may be severaltimesas greatasthestatisticitself.

000

28

APPENDIX 1[1

DEMOGRAPHIC TERMS

Age. —The age recorded for each person is the ageat last birthday. Age is recorded in single years.

Race. —Race is recorded as “White,” “Negro,” orr!Other. rJ !Iotherl t includes American Indian, Chinese,

Japanese, and so forth. Mexican persons are includedwith “White” unless definitely known to be Indian orother nonwhite race.

Population density .—The five classes comprisingthis characteristic were derived from the design of thesample which accomplished a stratification of the pri-mary sampling units by population density in each ofthree broad geographic locations. Because the Surveywas started in 1960, the primary sampling units withineach of the five population density classes were neces-sarily based on populations and definitions of the 1950census. The name of each selected primary samplingunit within each population density class and geo-graphic location, along with other selected sample data,are presented in an earlier report.z

The definitions for each of the five population densityclasses are as follows:

Giant metropolitan areas. — This class includesnine primary sampling units defined in the 1950 censusas a standard metropolitan statistical area (SMSA) andhaving a population of 3,000,000 persons or more.

OtheY veyy .laY.gemetropolitan areas. —Included inthis class are six standard metropolitan statistical areaswith a population of 500,000 to 3,000,000 as defined bythe 19S0 census.

Othev standard metropolitan statistical a?’eas.—This class includes nine other SMSA’S selected as pri-mary sampling units. With one exception— Providence,R.I.–all had less than 500,000 population.

Othe~ urban. —This includes eight primary sam-pling units which were highly urban in composition butwere not defined in 1950 as standard metropolitan areas.

Rural. -This includes 10 primary sampling unitswhich were primarily rural in composition according to1950 census definitions.

Region.—For the purpose of classifying the popu-lation by geographic area, the United States was dividedinto three major regions. This division was especiallymade for the design of the HES sample. The regions andthe States included are as follows:

Northeast -------- Maine, Vermont, New Hampshire,Massachusetts, Connecticut, RhcdeIsland, New York, New Jersey,Pennsylvania, Ohio, and Michigan.

South ------------ Delaware, Maryland, District ofColumbia, West Virginia, Virginia,North Carolina, South Carolina,Georgia, Florida, Keniucky,Tennessee, “Alabama, Mississippi,Arkansas, Louisiana, Oklahoma,and Texas.

West ------------ Washington, Oregon, California,Idaho, Nevada, Montana, Utah,Arizona, Wyoming, Colorado, NewMexico, North Dakota, South Dakota,Nebraska, Kansas, Minnesota, Iowa,Missouri, Wisconsin, Illinois, andIndiana.

Location of residence terms. —This term refers tourban or rural place of residence of the sample persons.For the first six primary sampling units at which exami-nations were conducted, the definition of urban and ruralwas the same as that used in the 1950 census. These lo-cations were Philadelphia, Pa., Valdosta, Ga., Akron,Ohio, Muskegon, Mich., Chicago, 111., and Butler, Mo.

For the remainder of the sampling units the 1960 censusdefinitions were used.

The change from 1950 to 1960 definitions is of smallconsequence in the Survey, since only six locations wereaffected, and the major difference is the designation in1960 of urban towns in New England and of urban town-ships in New Jersey and Pennsylvania.

According to the 1960 definition, the urban popu-lation comprises all persons living in(a) places of 2,500inhabitants or more incorporated as cities, boroughs ,vil-lages, and towns (except towns in New England, NewYork, and Wisconsin~ (b) the densely settled urbanfringe, whether incorporated or unincorporated, ofurbanized areas; (c) towns in New England and townshipsin New Jersey and Pennsylvania which contain no incor-porated municipalities as subdivisions and have either25,000 inhabitants or more or a population of 2,500-25,000 and a density of 1,500 persons or more per

29

square mile; (d) counties in States other than the NewEngland States. New Jersey, and Pennsylvania that haveno incorporated municipalities within their boundariesand have a density of 1,500 persons ormore per squaremile; and (e) unincorporated places of 2,500 inhabitantsor more not included in any urban hinge. The remaining~pulation is classified aa rural.

Size of place. —In this Survey the urban population isclassified as living “in the central city” or “outside thecentral city” of an SMSA. The remaining urban populationis classified as “not in SMSA.”

The definitions and titles of standard metropolitanstatistical areas are established by the U.S. Bureau ofthe Budget with the advice of the Fe +eral Committee onStandard Metropolitan Statistical Areas.

The definition of an individual standard metropolitanstatistical area involves two considerations: first, a cityor cities of specified population to constitute the centralcity and to identify the county in which it is located asthe central county; and, second, economic and social re-lationships with contiguous counties which are metro-politan in character so that the periphery of the specificmetropolitan area may be determined.

Persons “in the central city” of an SMSAare there-fore defined as those whose residency is in the city ap-pearing in the atand and metropolitan statistical areatitle. Persons residing in an SMSA but not in the cityappearing. in the SMSA title are considered to reside“outside the central city. ”

The remaining population is allocated into rural-farm and rural-nonfarm groups. The farm population in-cludes all persons living in rural territory on places of10 or more acres from which sales of farm productsamounted to $50 or more during the previous 12 monthsor on places of less than 10 acres from which sales offarm products amounted to $250 or more during the pre-ceding 12 months. Other persons living in rural territorywere classified as nonfarm. Persons were also clas-sified as nonfarm if their household paid rent for thehouse but their rent did not include any land used forfarming.

Employment status. —This term applies to the em-ployment status of persons during the 2-week periodprior to the week of interview. It is not intended that thisterm define the lalmr force or provide estimates of theemployed or unemployed population at the time of thesurvey.

Persons who reported that they either worked at orhad a job or business at any time during the 2-week per-iod prior to the week of interview were considered em-ployed, This includes paid work as an employee of some-one else, self-employment in business, farming, or pro-fessional practice, and unpaid work in a family businessor farm. Persons on layoff from a job and those whowere absent from their job or business because of tem-porary illness, vacation, strike, or bad weather are con-sidered as employed if they expected to work as soon asthe particular event causing their absence no longer ex-

isted. Free-lance workers are considered as currentlyemployed if they had a definite arrangement with one ormore employers to work for pay according to a weekly ormonthly schedule either full time or part time. Excludedare such persons who have no definite employmentschedule but work only when their services are needed.Also excluded are (1) persons receiving revenue from anenterprise in whose operation they do not participate,(2) persons doing housework or charity work for whichthey receive no pay, and(3) seasonal workers during theymtion of the year they were not working. (It should benoted that these data were not collected for Philadelphia.)

Occupation. –A person’s occupation maybe definedas his principal job or business. For the purposes of thisSurvey the principal job or business of a respondent isdefined in one of the following ways: If the person workedduring the 2-week-reference period of the interview orhad a job or business, the question concerning his occu-pation (or what kind of work he was doing) applies to hisjob during that period. If the respondent held more thanone job, the question is directed to the one at which hespent the most time. It refers to the one he considersmost important when equal time is spent at each job. Aperson who has not begun work at a new job, is lookingfor work, or is on layoff from work is questioned abouthis last full-time civilian job. A full-time job is definedas one at which the person spent 35 or more hours perweek and which lasted 2 consecutive weeks or more. Aperson who has a job to which he has not yet reportedand has never had a previous job or business is clas-sified as a “new worker. ”

The occupational groups are shown below with theappropriate census code categories.

Occupational title

Professional, technical, other kindredworkers, and managerial

Farmers and farm msnagersClerical and sales workersCrsftsmen, foremen, and kindred workersGperstives and kindred workersPrivate household and service workersFarm and other laborers (except mine)

Unknown(including new workers)

Census code

R,000-195, 250-285N, 222s, Y, z, 301-395Q, 401-545T, W, 601-721P, 801-803, 810-890u, v, x, 901,905,960-973995and all other codes

(U.S. Bureau of Census, 1960 Census of Population,Classified Index of Occupation and Industries, U.S. Gov-ernment Printing Office, Washington, D.C., 1960). This

information was not collected for Philadelphia andValdosta.

Education. —Each person is classified by educationin terms of the highest grade of school completed. Onlygrades completed in regular schools, where persons aregiven” a formal education, are included. A “regular”school is one which advances a person toward tin ele-mentary or high school diploma or a college, university,

30

or professional school degree. Thus, education invoca-tional, trade, or business schools outside the regularschool system is not counted in determining thehighestgrade of school completed.

Income of family OY unrelated individuals. -Eachmember of a family is classified according to the totalincome of the family ofwhich heis a member. Withinthe household all persons related to each other by blood,marriage, or adoption constitute a family. Unrelated in-dividuals are classified according to their own income.

The income recorded is the total of all income re-ceived by members of the family in the 12-month period

preceding the week of interview. Income from all sourcesis included, e.g., wages, salaries, rents from properties,pensions, help from relatives, and so forth.

Marital status.-lke categories of marital statusare mavried, m“dowed, divoyced, sepayated, and nevermawied. Persons with common-law marriages are con-sidered to be married. Separated refers to married per-sons who have a legal separation, those living apartwith intentions of obtaining a divorce, and other personspermanently or temporarily estranged from theirspouse because of marital discord.

——o oo—

31

* U.S.GOVERNMENT PRINTESGOFFICE :197?543-879/23

REPORTS FROM THE NATIONAL CENTER FOR HEALTH STATISTICS

Public Health Service Publication No. 1000

Series 1. Programs and collection procedures

No. 1. Origin, Progrmn, and Operation of the U.S. National Health Survey. 35 cents.

No. 2. Health Survey Procedure: Concepts, Questionnaire Development, and Definitions in the Health Interview Survey. 45 cents.

No. 3. Development and Maintenance ofa National Inven@ryof Hospi@ls and Institutions. 25 cents.

Series 2. Data evaluation and methods research

No. 1. Comparison of Two-Vision Testing Devicee. 30 cents.No. 2. Measurement of Personal Health Expenditures. 45 cents.

No. 3. The One-Hour Glucose Tolerance Test. 30 cents.No. 4. Comparison of Two Methods of Constructing Abridged Life Tables. 15 cents.

No. 5. An Index of Healtb: Mathematical Models.

Series 3. Analytical studies

No. 1. The Change in Mortality Trend in the United States. 35 cents.

No. 2. Recent Mortality Trends in Chile. 30 cents.

Series 4. Documents andcommittee reporte

Noreporte to date.

Series 10. Data From tbe Healtb Interview Survey

No. 1. AcuteConditions. Incidence and Associated Disability, United States, July 1961-June 1962. 40 cents

No. 2.No. 3.

No. 4.No. 5.‘i, 0. 6.

No. 7’.No. 8.

No. 9.

No. 10.No. 11.

No. 12.No. 13.

No. 14.No. 15.

No. 16.No. 17.

Family Income in Relation to Selected Health ~harac~ristics, United States. 40 cents.Length of Convalescence After Surgery, United States, July 1960-June 1961. 35 cents.

Disability Days, United States, July 1961-June 1962. 40 cents.

Current Estimates From the Health Interview Survey, United States, July 1962-June 1963. 35 cents.Impairments Due to Injury, by Class and Type of Accident, United States, July 1959-June 1961. 25 cents.

Disability Among Persons in the Labor Force, by Employment Status, United States, July 1961-June 1962. 40 cents.Types of Injuries, Incidence and Associated Disability, United States, July l957-June 1961. 35 cents.

Medical Care, Health Status, and Family Income, United States. 55 cents.

Acute Conditions, Incidence and Associated Disability, United States, July 1962-June 1963. 45 cents.Health Insurance Coverage, United States, July 1962-June 1963. 35 cents.

Bed Disability Among the Chronically Limited, United States, July 1957-June 1961. 45 cents.Current Estimates From the Health Interview Surt!ey, United States, July 1963-June 1964. 40 cents.

Illness, Disability, and Hospitalization Among Veterans, United States, July 1957 -June 1961. 35 cents.Acute Conditions, Incidence and Associated Disability, United States, July 1963-June 1964. 40 cents.

Health Insurance, Type oflnsuring Or@nization and Multiple Covemge, Unitd States, July 1962-June 1963.Cbmnic Conditions and Activity Limitations, United States, July 1961-June 1963.

Series 11. Data From the Health Examination Survey

No. I. Cvcle [of the Health Examination Survey: Smple and Response. United States, 1960-1962. 30 cents.

No. 2.

No. 3.No. 4.No. 5.No. 6.No.- 7.

No. 8.

No. 9.

Glucose Tolerance of Adults, United States, 1960-1962. 25 cents.

Binocular Visual Acuity of Adults, United States, 1960-1962. 25 cents.

Blood Pressure of Adults, by Age and Sex, United States, 1960-1962. 35 cents.Blood Pressure of Adults, by Race and Region, United States, 1960-1962. 25 cents.Heart Disease in Adults, United States, 1960-1962. 35 cents.

Selected Dental Findings in Adults, United States, 1960-1962. 30 cents.Weight, Height, and Selected Body Dimensions of Adults, United States, 1960-1962.

Findings on the Serologic Test for Syphilis in Adults, United States, 1960-1962.

Series 12. Data From the Health Records Survey

No reprts to date.

Series 20. Data on mortality

No reports t.o date

Series 21. Data on natality, marriage, and divorce

No. 1. Natality Statistics Analysis, United States, 1962. 45 cents.

No. 2. Demographic Characteristics of Persons Merried Between January 1955 and June 1958, United States. 35 cents.

Series 22. Data from tbeprogmmof sample survey srela~d~ vital records

No reports to date.