Miliary tuberculosis in BritishColumbiaIan G. Campbell, B.sc, m.b., b.s.,* Vancouver, B. C.

Summary: The case records werestudied of all 48 patients withmiliary tuberculosis registered bythe British Columbia Division ofTuberculosis Control between 1967and 1971. Two groups of patientswere found, one group of 30in whom the diagnosis was madeduring life and antituberculoustherapy instituted, and a secondgroup of 18 patients in whom thediagnosis was made only at autopsyand who had received noantituberculous therapy. In thelatter group there was a particularlyhigh incidence of concomitantdisease; additionally, tuberculintesting, bacteriological searchfor acid-fast bacilli and chestradiography were frequentlyneglected. This group consistedmainly of elderly male Caucasians.The disease was characteristicallyinsidious in onset, frequently withnon-specific symptoms andphysical findings. Fever was acommon presenting symptom.Several patients presented with afulminant illness which, althoughdiagnosed, was not affected byantituberculous therapy and was

rapidly fatal. North AmericanIndians showed a high incidenceof disease but a low mortalityrate, probably owing to theiryounger age.

In recent years many authors havestressed the changing pattern ofmiliary tuberculosis. Formerly a dis¬ease primarily of children, it is nowmost common in the eighth decade.1Diagnosis is frequently made onlyat autopsy.2-4 Miliary tuberculosis isfound in approximately 10% ofpatients with a fever of unknownorigin.5

Varied associated hematologicalreactions including leukopenia, pan-cytopenia, leukemoid reactions,agranulocytosis6 and monocytosis1have been described. A high in¬cidence of concomitant disease1'4,7further complicates diagnosis. Chestradiography is frequently negative8and tuberculin anergy is common.9The present report describes the

pattern of miliary tuberculosis inBritish Columbia over the five-yearperiod, 1967-1971.

Material

The case records of all 48 patientsregistered with the B.C. Division ofTuberculosis Control, Vancouver as

having miliary tuberculosis werestudied retrospectively. All cases ofmiliary tuberculosis occurring inBritish Columbia are registered, in¬cluding cases in which the diagnosisis made at autopsy. Since the crudeautopsy rate in British Columbiafor the year 1970 was only33.2/100 deaths it is quite possiblethat the actual number of patientsdying of miliary tuberculosis un¬

diagnosed during life may be con¬

siderably higher than the registered

figures would indicate. There wasno appreciable difference in thenumber of cases reported each year.The patients can be divided intotwo categories. The first groupcomprises 30 patients in whom thediagnosis was made during life andwho were treated for miliary tuber¬culosis. The second group consistsof the remaining 18 patients inwhom the diagnosis was reachedonly at the postmortem examinationand who had received no antituber¬culous therapy. Cases of tuberculousmeningitis with no other evidence ofmiliary spread were excluded fromthe present series.Results

Age and sex distribution in the 48cases are shown in Table I. In bothgroups of patients there were twiceas many males as females. In thediagnosed and treated group therewas fairly uniform age distribution,compared with the autopsy-diag-nosed- group in which 17 patients(94%) were over the age of 50.Racial originThis is shown in Table II. Of the26 Caucasian patients, 21 wereborn in Canada, two in Ireland,two in Portugal and one in Fin¬land. In Caucasians and Asiaticsover half of the cases occurred inindividuals over 60 while half ofthe North American Indians werebelow the age of 30. Over 80%of the autopsy-diagnosed groupwere Caucasians, whereas-43% ofthe diagnosed and treated group

?Medical Resident,Department of Medicine,University of British Columbia,Vancouver General HospitalReprint requests to: Dr. StefanGrzybowski, Faculty of Medicine,University of British Columbia,Vancou General Hospital, Vancouver 9,B.C.

C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108 1517

were North American Indians. In1970 there were just over 50,000Indians in British Columbia whilethe total non-Indian population was

2,135,000.10

Symptomatology

This was recorded in 42 of the 48patients. Non-specific symptoms in¬cluding general malaise and weak¬ness were prominent, being presentin 28 (67%) of the patients. Coughwas present in 26 (62%) and spu¬tum production in 17 (40%) of thepatients. Weight loss was also fre¬quent, occurring in 24 (57%). Lesscommon symptoms included nightsweats (four patients), headache(four patients) and urinary symp¬toms (three patients). Six of thetwelve autopsy-diagnosed cases inwhom symptoms were recorded hadno respiratory symptoms (50%), as

compared with 9 of the 29 patientsin the diagnosed and treated group(31%).

Physical findings

Detailed records were available for37 patients. Fever was the singlemost common finding, being pres¬ent in 28 (76%) of the patients.Crepitations of varied type and lo¬cation were observed in 20 (54%).Hepatomegaly was found in ninepatients, three of whom were alco¬holics; splenomegaly was present intwo and generalized lymphadeno¬pathy in, one patient. Seven of thepatients had impaired conscious¬ness, ranging from confusion tocoma, but meningism was observedin only one patient. Six patients(16%) presented with fever of un¬known origin and no other signifi¬cant symptomatology or physicalfindings; four were in the autopsy-diagnosed group. There were noother major differences between thetwo groups with respect to the phy¬sical findings.Concomitant disease

This was present in 23 of the pa¬tients, of whom 15 were in theautopsy-diagnosed group and 8 inthe diagnosed and treated group.

In the former, co-existing dis¬eases included atherosclerotic heartdisease (five patients, none of whomhad acute coronary thrombosis as

cause of death), dementia (two pa¬tients), pneumonia and chronicpyelonephritis (each two patients).Conditions affecting only a singlepatient included chronic myelogen¬ous leukemia, Wegener's granulo-matosis, carcinoma of the breastand pulmonary embolism. Two pa¬tients died from massive gastroin¬testinal bleeding, one each frompresumed "stress" ulceration of theduodenum and stomach.

In the diagnosed-and-treatedgroup the most common co-existingdisease was alcoholism with fattyinfiltration of the liver which was

present in three patients, one ofwhom died. Other concomitant ill¬nesses included systemic lupuserythematosus, silicosis, traumaticquadriplegia and atheroscleroticheart disease. One patient had anaortic valve homograft.

Although the incidence of con¬

comitant disease was high (83% inthe autopsy-diagnosed group), itwas considered to have contributed,directly or indirectly, to death inonly seven patients, all of whomwere in the autopsy-diagnosedgroup.

Bacteriology

The results of smear and cultureexaminations for Mycobacteriumtuberculosis are shown in Table III.Of the group of patients in whom

the diagnosis was made during life,positive smears were obtained intwo thirds and positive cultures inover 85% of the 30 cases. Smearsand cultures were negative in threepatients in this group.

In 16 of the 18 autopsy-diag¬nosed cases no smears or culturesfor Mycobacterium tuberculosis wereexamined antemortem. In one pa-

Table IIPatient racial origin and age incidence

?Expressed as a percentage of the total diagnosed-treated**Expressed as a percentage of the total autopsy-diagn<

Table IIIExamination for Mycobacterium tuberculosis

oup (30 patients)group (18 patients)

Smear CultureNumber Number positive Number positiveexamined andcultured

Diagnosismade duringlife (30)

Diagnosismade atautopsy

(18)Pleuralfluid

*A11 autopsy specimens**Two specimens obtained at autopsy

1518 C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108

tient in this group cerebrospinalfluid and pleural fluid were bothnegative for acid-fast bacilli on

smear, but positive cultures were ob¬tained after the patient died.

Radiology

Chest radiography was performedin 42 patients, of whom 30 werein the diagnosed and treated groupand 12 in the autopsy-diagnosedgroup. In the diagnosed and treatedgroup there were 28 patients whosechest radiograph showed a charac¬teristic miliary pattern of disease;

Table IVHematological findingsHemoglobin

(g./lOO ml.)MaleFemale

>14 12-14 10-12 105 8 4 50 4 4 2

Leukocytosis(>10,000/c. mm.)_9 (30%)

Polymorphonuclear leuko¬cytosis (>7500/c. mm.) 8 (27%)

Lymphocytosis03500/c. mm.)_1 ( 3%)

Monocytosis (>800/c. mm.) 2 ( 7%)Eosinophilia**

Q440/C. mm.)_1 (3%)Leukopenia (<4000/c. mm.) 2 ( 7%)Neutropenia (<1500/c. mm.) 2 ( 7%)Lymphopenia_(< 1500/c. mm.) 16 (53%)*High staff count

«400/c. mm.) 13 (43%)?Includes one patient with leukopenia.**Patient on prednisone.

Table VHepatic biochemistry

the remaining 2 patients had nega¬tive radiographs.

In the autopsy-diagnosed groupno chest radiograph was obtained insix patients. In another six patients,although radiography showed a

miliary pattern, a diagnosis of mili¬ary tuberculosis was either con¬sidered and rejected or not sus¬

pected. Two patients had entirelynormal chest radiographs and inthe remaining four cases non-mili-ary abnormalities were found (pleu¬ral effusion, basal pleural thicken¬ing, opacities in right upper lobeand diffuse nodular opacities).Tuberculin testingThis was carried out in 30 patients.Using 5 units PPD, 24 positiveresults were obtained, including twoin patients who did not react toinitial tuberculin testing.

Hematological investigationsHemoglobin and leukocyte countswere available for 30 cases andhemoglobin estimation alone in afurther two cases. Upper and lowerlimits for leukocyte count abnorma¬lities were taken from Dacie andLewis.11 Results are presented inTable IV. A mild to moderate ane¬mia was present in 72% of thepatients.The most common leukocyte ab¬

normality was lymphopenia whichwas present in 53% of cases. Poly¬morphonuclear leukocytosis was

Serum alkaline phosphatase > 35 IU/1.(Normal 9-35 IU/1.)Serum glutamic oxaloacetictransaminase >35 IU/1.(Normal 5-35 IU/1.)Serum lactic dehydrogenase > 350 IU/1.(Normal 150-350 IU/1.)

No. of patients No. of alcoholics

6 (12)*

5(12)

6(12)Serum total bilirubin > 1.0 mg./lOO ml(Normal < 1.0 mg./lOO ml.) 2(5)?Figures in parentheses = number of patients in whom test was performed

Table VIFatality and age of treated patientsAgeTotal

0-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90+

Dead

found in 27% of cases and was

usually associated with a high staffcell count. Other leukocytic ab¬normalities were relatively uncom¬mon. There were no significant dif¬ferences in leukocytic abnormalitiesbetween the group of diagnosed andtreated cases and those diagnosedat autopsy.

Hepatic biochemistryOne or more biochemical tests ofliver function were carried out in16 patients (Table V). Alkalinephosphatase, glutamic oxaloacetictransaminase and lactic dehydro¬genase levels were abnormally highin approximately 50% of testedpatients. Only one of the patientswith an elevated alkaline phospha¬tase level was an alcoholic.

Hepatomegaly was present infive patients who had one or moretests of hepatic biochemistry per¬formed, and abnormal results wereobtained in four of them. Of 11 pa¬tients without hepatomegaly, oneor more of the liver function testswas abnormal in six.

Outcome

Tables I and VI show the relation¬ship of fatality to age in the twogroups of patients. Of the 30 pa¬tients in the diagnosed and treatedgroup 23 recovered from their ill¬ness, remaining well at follow-up(from six months to five years) withno recrudescence of their disease.In this group four (44%) of thenine patients over the age of 60died while there were only threedeaths (14%) among the 21 pa¬tients under the age of 60. Of theseven patients who died, six pre¬sented with a fulminant illnesswhich, although correctly diagnosedand treated, led rapidly to deathusually within days after admissionto hospital. Six of the seven deathsin this group were in men. Of theautopsy-diagnosed group, 13 (72%)of the 18 patients were over theage of 60.

Autopsy material

Autopsies were performed on 23of the 25 patients. The lung wasinvolved in all cases, the spleen in16, liver in 14, kidney in 10 andlymph nodes in 9 cases. Brain,meninges and adrenal glands wereeach involved in five cases. The

C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108 1519

pancreas, pericardium, bone andprostate were each involved in twoor three cases.

Concomitant disease (see above)was either confirmed or newly diag¬nosed in 15 of the 23 cases whocame to postmortem examination.

Table VII gives, in summary, a

profile of the 18 autopsy-diagnosedpatients.Discussion

The present study shows a dis-concertingly high rate of failure indiagnosis of miliary tuberculosisduring life in British Columbia asin other areas.2'4'12'13 There are sev¬eral factors responsible for the fail¬ure of recognition. The most im¬portant is the insidious nature ofthe disease as it occurs in the eld¬erly. In the 1946 autopsy series ofChapman and Whorton 14 only 25%of cases were diagnosed ante mor¬tem although the average age waslower and tuberculosis still a com¬mon disease. The peak age inci¬dence of miliary tuberculosis hassteadily increased over the years;1'913in the present series it is only inthe autopsy-diagnosed group thata marked predominance of elderlypatients is found. The high inci¬dence of concomitant disease inthese patients is another importantfactor in the failure to diagnosemiliary tuberculosis.1'15

Seventeen of the dead patients inthe present series had other dis¬eases and 15 belonged to the autop¬sy-diagnosed group. Atheroscleroticheart disease and alcoholism were

particularly common. Corticosteroidtherapy appeared to be an im¬portant initiating or potentiatingfactor in four cases; three of thesepatients died without the diagnosisbeing made before death. Similarfindings have been reported byearlier workers.16'17 However, con¬comitant disease significantly con¬tributed to or directly caused onlyseven of the deaths, all in the autop¬sy-diagnosed group. It is interestingthat two patients died of massivegastrointestinal hemorrhage fromacute "stress" gastric and duodenalulcers.The common presenting symp¬

toms were general malaise, weak¬ness, weight loss and cough withsputum production, which appearedinsidiously over weeks or months;1522 C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108

this was also found by Biehl.9 How¬ever, a fulminant course was seenin six patients who, although re¬

ceiving the correct diagnosis andintensive treatment, died within afew days of admission to hospital.This fulminant course was not seenin the autopsy-diagnosed patients,who usually had an insidious, non-

specific illness, with noticeable ab¬sence of respiratory symptoms or

signs. Petersdorf and Beeson18 andBottiger5 described tuberculosis(usually miliary) in 11% of thepatients admitted for investigationof fever of unknown origin and thismode of presentation was found insix patients (13%) in the presentseries, four of whom were in theautopsy-diagnosed group.

Fever was the single most fre¬quent physical finding in our cases,being present in 76% of the pa¬tients. Crepitations were present in54%, a figure slightly lower thanthat found by Biehl.9 Nine patients(three of them alcoholics) had hepa¬tomegaly. Unlike Litten19 who foundsplenomegaly in 70% of cases, ourincidence was only 5%. Retinaltubercles, found by Carpenter andStephenson20 in 75% of cases ofmiliary tuberculosis, were only re¬corded in one patient. This prob¬ably represents failure of thoroughophthalmoscopic examination, theimportance of which has beenstressed by Illingworth and Wright21and Chapman and Whorton.14

Tuberculin anergy is often asso¬ciated with miliary tuberculosis.9 In

the present series only six of thethirty patients in whom tuberculintesting was carried out showednegative reactions to 5 units ofPPD. Of the 18 patients diagnosedat autopsy, tuberculin testing wascarried out in only three, furtheremphasizing the failure to considertuberculosis in an undiagnosed ill¬ness.

There was a high incidence ofleukocytic abnormalities, i.e. lym¬phopenia, polymorphonuclear leu¬kocytosis and a high staff count, inthis series. Other leukocytic ab¬normalities, e.g. agranulocytosisand leukemoid reactions,6 neutrope¬nia and moncytosis1 were not seen.The polymorphonuclear leukocytosisis unexpected and cannot be ex¬plained by concomitant bacterialinfection, which was definitelypresent in only two such patients.Oswald,6 however, described a

relatively common transient poly¬morphonuclear leukocytosis whichwas usually associated with a sec¬ondary bacterial infection, andChapman and Whorton14 describedpolymorphonuclear leukocytosis in89% of their cases. The latter work¬ers also found lymphopenia andhigh staff cell counts as commonleukocytic abnormalities. The mildto moderate degree of anemia seenin 72% of the patients is in agree¬ment with the findings of Chapmanand Whorton14 and Hegler.22 Ab¬normalities of liver biochemistrywere demonstrated in over half ofthe relatively small number of pa-

Table VIIProfile of the 18 cases discovered at autopsy

tients in whom one or more testsof liver function were performed.In agreement with the findings ofBrunner and Haemmerli7 the mostcommon abnormality was an elev¬ated serum alkaline phosphatase,but SGOT and lactic dehydrogenaseelevations were also seen in a sig¬nificant proportion of patients.The great majority of the patients

in the diagnosed and treated grouphad a classic miliary pattern on

chest radiography. In the autopsy-diagnosed group six patients hadcharacteristic miliary shadowing inchest radiographs which were in-correctly diagnosed, while six pa¬tients had radiological changesother than a miliary pattern or a

negative radiograph. A proportionof the autopsy-diagnosed patientsprobably represent the "cryptic"group of patients described byProudfoot et al,1 patients who havean insidious illness with the absenceof miliary mottling in the chestradiograph. One aspect of the con¬

cept of "cryptic" disease is sup¬ported by the work of Berger andSamortin8 who, in reviewing theliterature, found 99 normal initialchest radiographs in 290 patientswho eventually were proved to havemiliary tuberculosis.

Bacteriological confirmation ofthe diagnosis was obtained in 70%of patients and in seven cases thiscame from autopsy material. Inthe diagnosed and treated group,positive cultures were obtained in85% of the cases and there was a

surprisingly high incidence of posi¬tive smears (66%). The high yieldof positive cultures from sputum,urine, stomach washings and cere¬

brospinal fluid suggests that wide¬spread use of bone or liver biopsyis unnecessary. On the other hand,Brunner and Haemmerli7 report 10out of 12 positive liver biopsies incases with diagnostic difficulty; allthese patients were treated andsurvived as compared with a further12 cases in which no biopsy hadbeen performed and the diagnosiswas disclosed only at autopsy. Itis probable that some of the casesnot identified before death wouldhave been revealed by liver or bonebiopsy.28

North American Indians havebeen shown to have a high diseaserate but low mortality from miliarytuberculosis as compared to Cauca¬

sians. Tuberculosis is 10 times morecommon among Indians,24 and hasa particularly high incidence amongIndian children.25 The lower mor¬

tality rate in the Indian populationis probably due to the younger ageat which the disease occurs, andthe more obvious clinical presenta¬tion in the majority of cases whichleads to greater ease of diagnosis.A similar situation has been de¬scribed by Falk26 in Negro andCaucasian populations in the UnitedStates.The present report confirms a

disturbingly high autopsy diagnosisrate for miliary tuberculosis inBritish Columbia, in accordancewith other Western experience. In¬sidious onset, lack of diagnosticallysignificant physical findings, and a

high incidence of concomitant dis¬ease in an increasingly elderlypopulation are largely responsiblefor this situation.

I wish to thank Dr. Stefan Grzybowskifor considerable help and encourage¬ment during the preparation of thispaper, and Dr. David Mowat, Director

of the British Columbia Division ofTuberculosis Control, for permissionto study case records from his Depart¬ment.

ResumeLa tuberculose miliaireen Colombie britanniqueLa presente etude passe en revueles dossiers cliniques de 48 maladessouffrant de tuberculose miliaire en-

registres de 1967 a 1971 par laBritish Columbia Division of Tuber¬culosis Control. On a classifie lesmalades en deux groupes: un grou¬pe de 30 malades dont le diagnosticavait ete porte pendant leur vie etqui avaient regu une medicationantituberculeuse et un second grou¬pe de 18 malades dont la maladien'avait ete diagnostiquee qu'a l'au-topsie et qui n'avaient regu aucunemedication antituberculeuse. Dansce dernier groupe, on notait une

frequence particulierement eleveede pathologies concomitantes. Deplus, l'epreuve a la tuberculine, lesrecherches bacteriologiques de ba-cilles acido-resistants et les radio-

AMNIHOOK

sterile,disposable

amniotic membraneperforator

Each AmniHook is individually sterile-packedfor immediate use in the labor room. Therounded, blunt end with a protected point re-

duces the chances of injury to mother and fetus.Time and costs of resterilization are saved, sincethe AmniHook is discarded after use.

HOLLISTER LIMITED . 332 CONSUMERS RD., WILLOWDALE, ONT.

C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108 1523

PYOPEN*BRIEF PRESCRIBINGINFORMATION

CONTRAINDICATIONS: Carbenicillin shouldnot be used in patients with a history ofpenicillin or cephalosporin allergy. Carben-icillin is hydrolyzed by staphylococcal beta-lactamase and is, therefore, contraindicatedin infections caused by beta-lactamase-pro-ducing staphylococci.

PRECAUTIONS: The same as for penicillin G.Resistant organisms have arisen during treat-ment. Adequate dosage should therefore bemaintained and intermittent sensitivity testscarried out. With high doses of PYOPEN theconcomitant administration of sodium may behazardous in patients with serious cardiacdisease.One gm of PYOPEN contains 6.5 mEq ofsodium ion. Therefore, large doses of thedrug should be administered with caution topatients with congestive heart failure, severehypertension, or edema.

ADVERSE REACTIONS: Similar to those re-ported with penicillin G. Also reported: painand rashes at site of i.m. injection; thrombo-phlebitis after prolonged administration; pru-ritus; eosinophilia; nausea; occasional rise inSGOT and alkaline phosphatase levels; onecase of an unexplained drop In hemoglobinlevel; one case of seizure in a hemiplegicpatient on high doses and one case of in-creased muscle weakness in a patient withmyasthenia gravis have been reported.Prolonged bleeding time has been reportedIn 22 of a series of 30 patients receiving be-tween 500 and 750 mg/kilo daily of PYOPENduring 14 days (giving blood levels of 200-400 ug/ml). Aggregation of platelets by aden-osine diphosphate was decreased in all 30patients. The defect appeared within 12 hoursafter starting therapy and took from 3 to 7days to disappear after discontinuing thedrug.Two women receiving 30 gm of PYOPEN dailydeveloped hypokalemia after 7 and 9 daysrespectively. Potassium levels were 2 and1.5 mEq/l. Oral potassium chloride rapidlycorrected the electrolyte disorder.

DOSAGE: Relatively high doses of carben-icillin are required in the treatment of severePseudomonas infections.ADULTS - Severe and overwhelming infec-tions (septicemia, extensive burns and wounds,pneumonia, meningitis, peritonitis): 12 to 30gm daily intravenously or by infusion with orwithout 1 gm of probenecid orally 3 timesdaily.In moderately severe infections: 1 gm every4 hours for 5 to 10 days.CHILDREN - Dosage, according to weightand age and severity of infection, may varyfrom 100-300 mg/kg/body weight.

PYOPEN - the most exact therapyin severe and overwhelming pseu-domonas infections.

Complete prescribing information andreferences available on request

IAye IstAyerst Laboratories division ofAyerst, McKenna & Harrison,Limited, Montreal, Canada

^T.M. Reg'd.Made in Canada by arrangement withBEECHAM RE5EARCH LABORATORIES

graphies pulmonaires etaient sou-vent negligees. Ce groupe 6tait sur-tout compose d'hommes ages derace blanche. La maladie avaitcomme caracteristiques un debutinsidieux, une symptomatologie etdes signes cliniques manquant sou-vent de specificite. La fievre etaitun sympt6me courant. Plusieurs deces malades ont presente une ma-ladie foudroyante qui, bien quetraitee, n'a pas reagi au traitementantituberculeux et qui eut rapide-ment une issue fatale. La maladie,tres frequente chez les Amerindiens,n9'a ete frappee que d'une faiblemortalite, probablement en raisondu jeune age des malades.

References

1. PROUDFOOT AT, AKHTAR AJ, Dou-GLAS AC, et al: Miliary tuberculosisin adults. Br Med J II: 273, 1965

2. FULLERTON JM: Unsuspected miliarytuberculosis. Br Med J I: 715, 1969

3. JUSTIN-BESANCON L, CHRETIEN J,GRIVAUX M: Tuberculoses mortelleset tuberculoses meconnues. Conclu-sions generales. Sem Hop Paris 42:1125, 1966

4. TREIP C, MYERS D: Fatal tuberculo-sis in a general hospital. A diagnosticproblem. Lancet I: 164, 1959

5. BOTTIGER LE: Fever of unknownorigin. III. The final diagnosis inpatients admitted as cases of feverof unknown origin. Acta Med Scand156: 469, 1957

6. OSWALD NC: Acute tuberculosis andgranulocytic disorders. Br Med JII: 1489, 1963

7. BRUNNER K, HAEMMERLI UP:Needle biopsy of the liver in theearly diagnosis of miliary tubercu-losis. Ger Med Mon 9: 372, 1964

8. BERGER HW, SAMORTIN TG: Mili-ary tuberculosis: diagnostic methodswith emphasis on the chest radio-graph. Chest 58: 586, 1970

9. BIEHL JP: Miliary tuberculosis. Areview of 68 adult patients admittedto a municipal general hospital. AmRev Resp Dis 77: 605, 1958

10. ROWE J: Personal communication,1972.

11. DACIE JV, LEwIS SM: PracticalHaematology, fourth ed. London,Churchill, 1968, p 13

12. BOTTINGER LE, NORDENSTAM HH,WESTER PO: Disseminated tubercu-losis as a cause of fever of obscureorigin. Lancet I: 19, 1962

13. JACQUES J, SLOAN JM: The chang-ing pattern of miliary tuberculosis.Thorax 25: 237, 1970

14. CHAPMAN CB, WHORTON CM: Acutegeneralized miliary tuberculosis inadults. N Engi 1 Med 235: 239,1946

15. ASHBA JK, BOYCE JM: Undiagnosedtuberculosis in a general hospital.Chest 61: 447, 1972

1l6. DOERNER AA, NAEGELE CF, REGANFD), et al: The develovmn:Lrt of tu-

berculous meningitis following cor-tisone therapy. Am Rev Tuberc 64:564, 1951

17. HEATON J, MCCAUGHEY WTE:Milary tuberculosis in adults. Ir JMed Sci 140: 251, 1971

18. PETERSDORF RG, BEESON PB: Feverof unexplained origin. Report on100 cases. Medicine (Baltimore) 40:1, 1961

19. LIrrEN M: Uber akute Miliartuber-culose. Klin Vortrage (Volkmann)19: 1061, 1875

20. CARPENTER G, STEPHENSON S:Tuberculosis of the choroid. Bull SocFr Ophthalmol 23: 288, 1906

21. ILLINGWORTH RS, WRIGHT T: Tuber-cles of the choroid. Br Med J II:365, 1948

22. HEGLER C: Akute allgemeine mili-artuberkulose, in Handbuch derTuberkulose, vol 4, third ed, editedby BRANER L, SCHROEDER G, BLUM-ERFIELD F, Leipzig, J Barth, 1923,p 1

23. HEINLE EW JR, DENSEN WN, WEs-TERMAN MP: Diagnostic usefulnessof the marrow biopsy in dissemi-nated tuberculosis. Am Rev RespDis 91: 701, 1965

24. Royal Comission on Health Services.Ottawa, Queen's Printer, 1965

25. GALBRAITH JD, GRzyBowsKi S, LAwCL, et al:: Tuberculosis in Indianchildren: primary pulmonary tuber-culosis. Can Med Assoc J 100:497, 1969

26. FALK Az U.S. Veterans Administra-tion - Armed Forces co-operativestudy on the chemotherapy of tuber-culosis. XII. Results of treatment inmiliary tuberculosis: a follow-upstudy of 570 adult patients. AmRev Resp Dis 91: 6, 1965

FORTHCOMINGMEETINGSCANADA

ANNUAL SCIENTIFIC FALL MEETING, ONTARIOMEDICAL ASSOCIATION, ANAESTHETIC SECTION,ONTARIO DIVISION. Sarnia, Ont. September 20t22,1973. Information: Dr. K. Panwar, Box 2222,Sarnia, Ont. N7T 717

ANNUAL MEETING, ONTARIO SOCIETY OF MEDICALTECHNOLOGISTS. Ottawa. September 27-28, 1973.Information: Dr. Marie Aprile, Dept. of Surgery,Banting Institute (Room 322), 100 College St.,Toronto 5, Ont.

OTHER COUNTRIES

9Th INTERNATIONAL CONGRESS OF BIOCHEMISTRY.Stockholm. June 28-July 4, 1973. Information:Mrs. G. Aulin-Erdtman, Swedish Society of Chem-istry, Wenner-Gren Center VI, S-113 46 Stock-holm, Sweden.

INTERNATIONAL SYMPOSIUM ON THE IDEWlFICArTION OF TABLETS. Pont A Mousson, France. June30-July 1, 1973. Information: J.-F. Lorentz, CentreAnti-Poison, Centre Hospitalier Regional, 54, Nancy,FranceTHCE HANS BERGEN CENTENARAY SYMPOSIUM ONEPILEPSY. Edinburgh, ScotlaSnd. July 1-5, 1973.Theme: The Natural H-istory and Management ofEpilepsy. Information: Kurt Fleischmann, Sym-posium Secretariat, Kurt Fleischmann and Asso-ciates, 164 North Gower St., London, NW1 2ND.

VIIIii INTERNATIONAL CONFERENCE ON HEALTHEDUCATIONt. Paris. July 8-14, 1973. Information:Secretariat, Vl Ith International Conference onHealth Education, 20 rue Greuze, 75 Paris 16e,France

1526 C.M.A. JOURNAL/JUNE 23, 1973/VOL. 108