metastases were found the liver scanwas normal, and in three of the seventhe serum LDH value was raised (80,87 and 290 lU/i). In all 12 of thepatients in whom liver metastaseswere found at autopsy the serumLDH value was elevated, but in 4 ofthe 12 the liver scan made at thetime of the elevated serum LDHvalue was considered to be normal.

Discussion and conclusion

We believe that this study hasprovided confirmation of the findingsof Einhorn and colleagues,1 havingdemonstrated the usefulness of anelevated serum LDH value as an in-dicator of hepatic metastases in pa-tients with stage ITIB or IV malignantmelanoma.. If a serum LDH value

greater than 78 lU/i had been con-sidered an indication for requestinga liver scan, 60 instead of 159 liverscans would have been done; 3 ab-normal scans and 1 equivocal scanwould have been missed. Since we donot have pathologic confirmation ofthe proportion of this group of pa-tients who had metastases, it is notstrictly correct to speak of test sensi-tivity or specificity. However, ourresults do suggest that measuring theserum LDH concentration may beconsidered as a prerequisite for order-ing a liver scan: if the LDH value isnormal the probability that the liverscan will be abnormal or equivocalis only 0.04 (4/99), while if the LDHvalue is elevated the probability thatthe liver scan will be normal is 0.68(41/60).

References

1. EINHORN LH, BURGESS MA, VALLEJOSC, et al: Prognostic correlations andresponse to treatment in advancedmetastatic melanoma. Cancer Res 34:1995, 1974

2. CASTAGNA J, BENFIELD JR, YAMADAH, et al: The reliability of liver scansand function tests in detecting meta-stases. Surg Gynecol Obstet 134: 463,1972

3. MCCREADY VR: Scintigraphic studiesof space-occupying liver disease. SeminNuci Med 2: 108, 1972

4. HOLDER LE, SAENGER EL: The use ofnuclear medicine in evaluating liverdisease. Sernin Roenigenol 10: 215,1975

5. DRUM DE, BEARD JM: Scintigraphiccriteria for hepatic metastases fromcancer of colon and breast. J NuciMed 17: 677, 1976

Severity of lung disease in Indian childrenC. STUART HOUSTON, MD; ROBERT L. WEILER,* B Sc; BRIAN F. HABBICK, MB, CH B

Recurrent bronchopulmonary infec-tions in North American Indian chil-dren are an important social andpublic health problem in Saskat-chewan and adjacent provinces. Wehave been impressed by the pre-valence and severity of these infec-tions, and feel that particular em-phasis should be given to measuresdesigned to prevent them. The studydescribed in this paper was conductedto investigate the problem.

MethodsBeginning in January 1974 a 3-

year prospective record was kept ofchildren in whom the roentgeno-graphic appearance of the chest be-came worse during treatment ofpneumonia in hospital. One of us

From the departments of radiology,pediatrics and pathology, UniversityHospital, Saskatoon*Fifth..year medical student

Reprint requests to: Dr. C. Stuart Houston,Department of radiology, UniversityHospital, Saskatoon, Sask. S7N 0W8

(R.L.W.) then reviewed the charts ofall children 14 years of age or lesswho had been admitted to our hos-pital with pneumonia during the sameperiod. Children with underlying con-ditions such as heart disease andcystic fibrosis were excluded.

Results

Indian children with pneumoniatended to be sicker. Fever and ad-ventitious sounds persisted more thantwice as long as with white children(Table I). The stay in hospital wasmuch longer for Indian children, andthis could be explained only in smallpart by differing social and distancefactors. Five times as many Indianshad associated diarrhea. Few Indianshad been breast-fed.Of the 21 children in whom the

roentgenographic appearance of thechest worsened in hospital, 20 wereIndians and 1 was a 12-month-oldwhite boy in whom adenovirus infec-tion developed after rubeola (TableII). None of the 64 white childrendied, but 4 of the 102 Indian children

died, 2 with pneumonia followingpertussis, 1 with rubeola and giant-cell pneumonia, and 1 with fulminat-ing primary staphylococcal pneumo-nia complicated by empyema.

Autopsy of patient 9, who died ofadenovirus pneumonia following per-tussis at 8 months of age, showed

Table I-Data for 166 chIldren with post'-monla, 1974-76

Indian WhiteVariable (n=102)(n=64)

Malefinale ratIo 59:43 32:32No. of patients withWorsening of mont-gonographic appear-ance of chest Inhospital 20 1

4 0Oeath In hospital 4 0

Average duration (d) ofStay I. 1 26.9 8.7

6.6 31Adventitious sounds 17.2 7.6.Wsesociatwi -diarrhea 48% 9%Breast-fed I month or

11174 91404Not recorded for some patients.

1116 CMA JOURNAL/MAY 5, 1979/VOL. 120

much more thickening of the bron-chial walls than would have been pre-dicted from the roentgenograms madejust before death.

DiscussionIndian children, when admitted to

hospital with other problems, com-monly have thick bronchial walls; theroentgenographic appearance is simi-lar to that of white children withasthma, cystic fibrosis or immune de-ficiency. When rubeola or pertussisdevelops in Indian children, broncho-pneumonia almost routinely follows.

These chest infections cause manydays of sickness, are a great concernto the parents, and often result intransportation of the child to a dis-tant city hospital in a different cul-ture, where he or she is treated bystrangers who speak a different lan-guage.The average Indian child is ad-

mitted to hospital five times as oftenas the average white child. In Sas-katchewan, published figures for In-dians are available only for 1966,when there were 1148 admissionsand 12 387 days in hospital per 1000Indian infants in their first year oflife, compared with only 258 admis-sions and 2305 days in hospital per1000 infants in the overall popula-tion, including Indians.1A chronic cough and increased

susceptibility to chest infections havebeen among the hallmarks of theunderprivileged through the ages. Anincreased incidence of severe pneu-monia is still found among the Maoriand Polynesian children in New Zea-land,2 and among the Aborigines ofAustralia,3 who also show bronchialwall thickening (W.J.H. Caldicott:personal communication, 1977).

In 1965 Brody4 reported an ap-parently increasing incidence of se-vere lower respiratory tract diseasein young Inuit children in Alaska. Of318 newborns followed prospectively35% had at least one episode ofpneumonia or bronchitis during thefirst year of life and 16 died fromrespiratory causes. A follow-up study

the upper lobe of the right lung, fol-lowing severe lower respiratory tractinfections in childhood, is unusuallycommon among Inuit and Indianchildren in Alaska.8

Herbert and colleagues7 studiedprospectively 97 Indian children with112 episodes of pneumonia that pre-sented to the Charles Camsell Hos-pital, Edmonton during 1 year. Ofthese, 71 % were between 3 and 18months of age. Two had a historyof more than six previous attacks and34 had records of three to six pre-vious attacks. Of the 14 children whoacquired adenovirus type 3 pneu-monia while in hospital with unre-lated problems 11 had a chronic pro-ductive cough 10 years later.8 Bron-chograms were done in seven andshowed widespread airway damage.In our series of children adenovirusinfections were present in the one

white child and in three of the In-dian children in whom the roentgeno-graphic appearance of the chestworsened in hospital.

Chest infections are commonamong treaty Indians, all of whomhave some white blood, and amongnontreaty Indians and M6tis who livein a similar economic and culturalsetting. Several factors contribute tothis. The average number of personsper house is 6.67 on SaskatchewanIndian reserves, compared with 345for non-Indians. Only 1.4% of In-dian homes on Saskatchewan reserveshave running water.9 In a small homethe children lack sleep because theystay awake until the adults go to bedand then have to get up early forschool. Some Indian homes are in-adequately or irregularly heated, andsome small homes are smoke-filledfrom the heating system or from ciga-

Table ll-.-.Details of patients in whom the roan raphic appearance of the chest worsened Inhospital

InfectionPatient Ageno. (mo) Sex Status4' Primary Complicating Outcome

1 3 F T

2 4 F T

$ 4 F T

4 4 F TP 5 F NT* 6 N T7 6 N T0 7 N t* 8 5 NT10 8 U T.

11 9 U T

12 11 N T

1$ 11 U T

141515

1212

14

U

U

U

NT

w

Pertussis

Influenza A/Vlctorl

Staphuk,coccusaureus

RubeolaPertusols

PertussisPertusais$hiqgUePartusslsRubeola

Rubeola

Rubeola

Rubuola

tRubeola

Rubeola

by Maynard and colleagues3 in thesame area in the winter of 1966showed a viral association in 75%of attacks. Bronchiectasis, often in

a

1

.apticmonk

Paraintl.. ewe.naumo.denovlWenovi

.neumo

.danovi

Hewjplrilusinfluenzue

CMA JOURNAL/MAY 5, 1979/VOL. 120 1119

rettes. The diet may be inadequate,especially relative to physical ac-tivity,10 and 35% of Indian childrenunder 4 years of age have low serumconcentrations of vitamin A.1'The change from breast-feeding to

bottle-feeding is both a social and abiologic problem. The need for theprotection offered by the substancesin breast milk is even greater in in-fants with an increased susceptibilityto disease.'2 Schaefer'3 has foundthat, among the Inuit, breast-fed chil-dren have much less otitis media thanbottle-fed children, and Ellestad-Sayed and colleagues'4 have reportedthat breast-fed Indian children innorthern Manitoba have far fewerlower respiratory and gastrointestinaltract infections than bottle-fed in-fants, even though the social condi-tions of the breast-fed infants are, ingeneral, poorer.

The high frequency with whichchest infections were associated withdiarrhea in the Indian children inour study is of interest. It may in-dicate a higher susceptibility of boththe respiratory and the gastrointes-tinal systems to infection, or it couldrelate to a common precipitatingcause, such as milk allergy, whichappears to be more frequent in In-dian infants than in white infants.15

Indian children have either normalor, more often, elevated concentra-tions of serum globulins; the eleva-tion of serum IgA, IgD, IgE and IgGconcentrations has been attributed toincreased exposure to infectionsrather than to genetic factors.16 Thissuggests a satisfactory humoral im-mune response; however, Indian chil-dren may have a different immuneresponse to some infections, com-pared with non-Indians. Rubeola wasabsent among Indians prior to settle-ment by the white man: such a highproportion of a sparse populationwould be killed or immunized thatthe virus could no longer propagate.17The lack of many centuries of expo-sure and selection could, in part, ex-plain the susceptibility of Indians notonly to the tubercle bacillus but alsoto specific viruses, including thatcausing rubeola, and adenoviruses.Rubeola is known to be more severein malnourished populations and, in

turn, it may precipitate malnutrition.18Sixteen of our 21 patients in whom

the roentgenographic appearance ofthe chest deteriorated in hospital, in-cluding 3 of the 4 that died, had avaccine-preventable disease - ru-beola or pertussis - before contract-ing pneumonia. However, of the 10such children with rubeola, 6 ac-quired it before 12 months of age,the age usually advocated for measlesvaccination. An active rubeola vac-cination program for Saskatchewantreaty Indians is now being offeredat the age of 7 months. It is realizedthat further immunization will benecessary at a later age, but the pro-tection from the devastating effect ofmeasles on young Indian childrenmakes early vaccination justifiable.

It is probable that recurrent pul-monary infections in Indian childrenmight largely be prevented by im-provement of living conditions, en-couragement of breast milk as thesole food for the first 6 months oflife, and more widespread early vac-cination.We are indebted to many people. Drs.Frank Scott, of Loon Lake, Sask., andStan C. Houston, of Ile-.-Ia-Crosse,Sask., provided criticism from the gen-eral practitioner's viewpoint, Drs. JohnW. Gerrard, of Saskatoon, and MaryEllen Avery, of Boston, did so from theacademic pediatrician's viewpoint andErnest Tootoosis, of Saskatoon, did sofrom the Indian's viewpoint; Dr. PatPrestage, of Regina, and Dr. V.L.Matthews, of Saskatoon, provided sta-tistical information; and Mrs. JoanMatlock typed many versions of themanuscript.

References1. Annual Report ol the Saskatchewan

Hospital Services Plan, 1966, Supple-ment II - Indian Hospital Care inSaskatchewan, Dept of Public Health,Saskatchewan, Regina, 1966, pp 122-36

2. LANG WR, HOWDEN CW, LAWS J, etal: Bronchopneumonia with serioussequelne in children with evidence ofadenovirus type 21 infection. Br MedJ 1: 73, 1969

3. MAXWELL GM: Chronic chest diseasein Australian aboriginal children. ArchDis Child 47: 897, 1972

4. BRODY JA: Lower respiratory illnessamong Alaskan Eskimo children. ArchEnviron Health 11: 620, 1965

5. MAYNARD JE, FELTZ ET, WULFE H,et al: Surveillance of respiratory virusinfections among Alaskan Eskimochildren. JAMA 200: 927, 1967

6. FLESHMAN JK, WILSON JF, COHEN H:Bronchiectasis in Alaska native chil-dren. Arch Environ Health 17: 517,1968

7. HERBERT FA, MAHON WA, WILKINSOND, et al: Pneumonia in Indian andEskimo infants and children: Part I.A clinical study. Can Med Assoc J96: 257, 1967

8. HERBERT FA, WILKINSON D, BURCHAKE, et al: Adenovirus type 3 pneumoniacausing lung damage in childhood.Can Med Assoc J 116: 274, 1977

9. Saskatchewan Indian Community En-vironmental Profiles 1977, Dept ofNational Health and Welfare, medicalservices branch, Ottawa, 1977

10. BROWN RE: Interaction of nutritionand infection in clinical practice. Pe-diatr Clin North Am 24: 241, 1977

11. Nutrition Canada: The Indian SurveyReport. A Report from Nutrition Can-ada by the Bureau of NutritionalSciences, Dept of National Health andWelfare, Ottawa, 1975, pp 73-83

12. GRULEE CG, SANFORD HN, HERRONPH: Breast and artificial feeding. In-fluence on morbidity and mortality of20000 infants. JAMA 103: 735, 1934

13. SCHAEFER 0: Otitis media and bottle-feeding. An epidemiological study ofinfant feeding habits and incidence ofrecurrent and chronic middle ear dis-ease in Canadian Eskimos. Can JPublic Health 62: 478, 1971

14. ELLESTAD-SAYED J, CooDIN FJ, DIL-LING LA, et al: Breast-feeding protectsagainst infection in Indian infants. CanMed Assoc J 120: 295, 1979

15. UPADHYAY YN, GERRARD JW: Recur-rent pneumonia in Indian children.Ann Allergy 27: 218, 1969

16. GERRARD JW, Ko CG, DALGLEISH R,et al: Immunoglobulin levels in whiteand Mdtis communities in Saskat-chewan. Clin Exp immunol 29: 447,1977

17. BLACK FL: Infectious diseases in pri-mitive societies. Science 187: 515,1975

18. MORLEY DC: Measles in the devel-oping world. Proc R Soc Med 67:1112, 1974

BOOKScontinued from page 1111PROGRESS IN CANCER RESEARCH ANDTHERAPY. Volume 11. Lung Cancer: Pro-gress in Therapeutic Research. Edited byFranco M. Muggia and Marcel Roz-encweig. 614 pp. Illust. Raven Press,New York, 1979. 545. ISBN 0-89004-223-3RECENT ADVANCES IN CLINICAL NEU-ROLOGY. Number Two. Edited by W.B.Matthews and Gilbert H. Glaser. 199 pp.Illust. Churchill Livingstone, Edinburgh;Longman Canada Limited, Don Mills,Ont., 1978. $30. ISBN 0-443-01793-X

A REVIEW OF BIOSTATISTICS. A Programfor Self-Instruction. 2nd ed. Paul E.Leaverton. 92 pp. Illust. Little, Brown andCompany (Inc.), Boston, 1978. $6.50,spiralbound. ISBN 0-316-51852-2

CMA JOURNAL/MAY 5, 1979/VOL. 120 1121