Document e5YE65wKOq7QQL61N8Lk6BwkE

November ip# 193& Mr* Roger S. Hubbard The Buffalo CcsK.ral Hospital Buffalo Hew York . .- ly dear Mr* Hubbard** Please accept my thanks for jour lofeter giving Information ov> jour analysas of urine samples fro- and discussing the significance of the reeuxts# j. mould like to give you cur results j, to comment on their sign ifleone and then to discuss the general problem of methods of analysis and methods of obtaining samples* Chi July 28th ne received from Dr* Killenbrsnd a sample of blood and a sample of approximately 100 co* of urine# both having been collected in our containers and according to our instructions# copies of which are enclosed* The result on the blood (0*535 mg* per 100 grans of hole blood) was quite within the nonaal limita# but that on the urina {Q#9l?7 ag* in a sample of 100 cc, equivalent to 0*1^7 mg* per liter) was not only at variance with the blood finding but also was so high aa oca-, coly to be bona fide* Despite the pre scribed, precautions# samples are sometimes contaminated# and when collected ?n industrial plants they are not uncommonly so* It is our custom# thorofore# In all cases in which the initial results are quite high# and especially hen the blood and urine differ materially in their signi ficance# to ciock them by obtaining additional samples, especially a sample of urine of large volume* In this instance this procedure resulted!In the following datai a sample of urine of 3I4.80 cc* contained load to the extent of 0*02 ng* per liter# while a sample of feces contained 0*08 mg* of lead or 0*06 mg* per gram of ash* These results justified the conclusion that the high result on the previous small sample of urine was due to contamination of the sample, and w so reported our conclusion to Dr* Elllenbrand* Subsequently shea further question was raised we obtained additional samples of blood# urine and feces# these giving results aa follows* 0*025 and 0*03 mg* per 100 grams of whole blood#0*05 rag* per liter of urine in a sample of 3180 cc*#and 0*08 mg* in a sample of focos# equivalent to 0*09 mg* per gram of feoal ash* The combined results in my opinion remove all doubt ns to the general level of lead excretion of this man# putting it within normallimits* The magnitude of the load exposure# as Indicated by these results# is In koeping with 2 4 the facts as to the occupational history of this man, and considering the atypical character of his clinical picture with respeot to lead, I believe lead can be excluded as an etiological factor* As to methods of analysis employed b y us, I am sending you separately the papers of Cholak and Hubbard from this laboratory. For a considerable period of time, and until we could satisfy ourselves as to the sensitivity and accuracy of the spectrographic method developed b y Cholak and the dithizone method as adapted by Hubbard, we relied upon a chemical method developed by Edgar, Thsmann and Gholak of this laboratory and described in the Journal of Industrial Hygiene for September 1933* It contains several modifications of Fairhall*s method which served to make it more accurate and uniform in Its application as well a3 specific with reference to lead. Fairhall criticized our modifications largely because he was offended by our comments on his method, but the fact is that he had never described his own improvements of the original method in such detail as to permit precise reproduction of his procedures, and therefore he had little justification for his irritation at us* That irritation has 3ince been dispelled, I believe, now that he and his associates in the Public Health Service are using both a dltiilzone method and a spectrographic method similar to Cholak*a, and are finding them quite satisfactory and far superior to the chromate method* *V recognized years ago that the chromate method was not adequate for handling blood and small samples of urine, and it was primarily for that reason that we were compelled to develop more sensitive methods for carrying on certain phases of our experimental work. You are quite right in your feeling that the chief criticism leveled at the chromate method referred to its loss of lead and its consequent low results* nevertheless we found it impossible to avoid occasional high results in urine samples, and frequent high results on feces due to the presence of other metals, without the incorporation of a sulphate step based on the work of the Australians, Avery, Hemmingway et all, as well as a special procedure when bismuth occurred. Hot only are significant errors associated with the use of the chromate method as described by Fairhall, under the most favorable circumstancesbut gross errors can bo introduced, and in our experience are introduced, b y the slight variations in technique which arise from infrequent and irregular performance of analyses, and by variations in the analytical environment and in reagents employed. The la3t two factors may result in gross errors. A plumber repairs a pipe line and the samples handled in the laboratory for that day or for several days yield high results. A dust is raised in cleaning or in moving objects in the laboratory and the analytical results go wrong. The glassware is handled with less than the usual precision and errors arise. By reason of sad experience ns have found It necessary to eliminate one factor after another and now no change nothing tn our analytical laboratories without stopping the work and cleaning thoroughly thereafter# Tie have had to exclude outside soot by filtering the air supply of our laboratories and we ha\e had to discontinue smoking In 9ueh laboratories# Dust la avoided with the utmost care# Glassware Is vjashed in accordance with a specially p r o scribed technique and C# P* bichromate Is used In the cleaning solution instead of the customary technical grade# Ttoagenba ere also specially purified to a known purity as doterm* nod by spectrograph! analysis# The dangers of error associated with analytical procedures in the hands of even a careful and experienced nan are very great but they are as nothing compared with the opportunities for error in the collection of samples# I am sure you would bo utterly astonished at the oppor tunities fc&* contamination which we have discovered in some fifteen years of work with thousands of samples but I shall not bor you with these details# Suffice it to 3ay that the analysist Is westing his time and misleading his clinical colleagues if he does not provide chemically clean containers and dotslied instructions for the avoidance of contamination in the collection of samples# nothing that ?q have been able to do in hospital practice has enabled us to escape occasional gross contamination of samples of urine# Blood camples token with hospital syringes and needles cleaned and sterilised in the usual cay hill always be seriously eontceilna tod* C'fe us a specially cleaned pyrex tube# supplying with it a special all-stainless steel . needle sterilised by boiling in triple-distilled water and encased in a second correspondingly leaned pyrex tub ready for us#) Urine samples collected in hospital urlnsla and poured into a bottle as in she case of ordinary 2 k hour samples will usuaXV be grossly contaminated# Here oar experience onabloo us to say that results In th urine over 0 , 3 0 mg* per liter are probably the results of contami nation since In fch most serious cases of load, exposure and intoxication on rarely finds results In excess of this level# To have found a vory fe results as high as 0#5P mg# per liter mostly in children b u t figures above this range do not occur except as th result of massive closes associated with acute and usually fatal Intoxication* Bosults of this order of magnitude o r greater are fch offoot of contamination of the sample cither in its collection or analysis# X have no hesitation therefore in saying that fch results obtained in your laboratory on August 2 9 th and September 2l*fcfa are fantastically and impossibly high as was also eur result of o#l*7 mg* per liter in a 100 ec# sample collected at the plant# Such concentrations simply do not occur except under the most unusual circumstances* Even your later samples however yielded high results# Our samples of blood on October 21st yielded 0#025 and 0#Q3 mg# of load per 100 grams end our urine sample on October 24th yielded only 0*05 rag* per liter* as contracted with your result of 0*12 rag* per liter on the same date* This is in spite of the fact that our analytical methods are more accurate and more sensitive than yours at its best* l*e there la little or no Inherent loss associated w i t h the application of cur methods* There must have been some factor of contami nation therefore, or some failure on the part of the analytical procedures #i X-.have got to some length to discuss the details of this case* but X should be glad to consider further any other points which may be raised* By reason of our experimental work and its Implications* Z have been, much more interested in such eases than clinicians in general would be, and moreover* I ora greatly disturbed by the Increasing reliance on certain laboratory data* the significance of which are not adequately appreciated by our clinical colleagues* I should also be glad if you would send us the aliquot of the sample you have analysed* We will analyse it spectrographieally and b y the di this one method* Incidentally w will be glad to dotera: the lead* copper* aluminum and silver quantitatively* and bismuth as well if it is found to be present under our conditions of combustion* Cordially yours* RAK-is hob'erfA* Echo*'.8. HZ' 0003591