in Albany in the spring of 1917. Although the case was investigated its source was not determined. Later in the year, Dr. Sears, the sanitary supervisor of the Syracuse district, investigated a case in that district and learned that a girl belonging to the household had recently returned from Albany. She had had typhoid about 18 months. before and it was while she was a member of Mrs. B's household that Mrs. B. was taken ill. An examination of her stools proved her to be a carrier. It is probable that such instances as Mrs. B's are not uncommon. Either the typhoid histories of other members of households are not obtained or their significance is overlooked.

The health officer should also have a knowledge of diagnosis, including an appreciation of laboratory aids to diagnosis, and should recognize not only their value but their limitations. He should thoroughly appreciate the fact that there are very mild atypical cases and that there are atypical cases which are very severe. Mild smallpox and mild scarlet fever are very common and the latter particularly is frequently difficult to diagnose. Mild typhoid and mild diphtheria are not infrequent and it is in these. cases that the laboratory is most useful. In the clinical cases-the "typical" cases the laboratory evidence is either simply confirmatory if positive or a warning to review critically the history, signs, and symptoms upon which the clinical diagnosis is based. There is too much of a tendency on the part of the profession to let the laboratory make the diagnosis-to assume that its findings are infallible-and this attitude is fraught with quite as much danger as the opposite one of disregarding the laboratory al

together. Antitoxin is not given because the culture was negative; an incipient phthisis is neglected because no bacilli are found in the sputum. On the other hand, a pelvic abscess, a pyelitis or a tuberculosis may be overlooked because of a positive Widal; the history and the clinical evidence being entirely subordinated to the microscope and test tube. This is not the fault of the laboratory but of the clinician. The latter does not go far enough with his physical examination or with the submission of material to the laboratory. Blood counts and urinalyses still occupy a place in nosology and every patient who is sick enough to have a Widal is entitled to a reasonably complete physical examination. A wholesome mixture of sanity is needed with our science.

The foregoing paragraphs merely review in a superficial way some of the "high spots" in the professional knowledge the health officer needs. He also needs-absolutely must. have, if he is to control the acute communicable communicable diseases-information concerning the patient, his environment, his associates, and his family. Objections may be raised to the health officer's obtaining this information-that it takes too much time, that the families or the attending physicians object to it, etc.; but if such objections are heeded. it simply means partial and ineffective efforts to control. To isolate "A," who has been reported as a case of scarlet fever, is no assurance that "B" and "C" will not contract the disease from the same source. We must find the unknown quantity X who or which infected. "A" and assure ourselves that X is innocuous. When all the others who may have been infected through X shall have been dis

covered and properly cared for, we may feel reasonably sure that further cases will not occur.

The determination of the source of infection is sometimes easy, often difficult, and frequently impossible, but it is always worth at good hard try. It is much more important than terminal fumigation, or concurrent disinfection, or even a strictly maintained quarantine. Every large outbreak is first a small one and every small one may be traced back to a single case or carrier. Though it may be less difficult to locate the sources of infection of half a dozen cases than of one (provided the source is common to all) yet it is correspondingly more difficult to control the half dozen and their contacts.

The investigation of cases for the purpose of discovering the source. of infection is perhaps the least understood procedure in the control

of communicable diseases. Aside from the fault that they are too frequently not even attempted the most frequent faults are that

I There is no written record made at the time of investigation; 2 The investigation or the record or both are incomplete;

3 The investigator begins with a preconceived notion of the source and bends statements to fit the theory;

4 Information is inaccurate;

5 The information obtained is not tabulated and studied from a statistical standpoint.

Unless a written record is made at the time of investigation-putting down each answer as it is given the investigation and its resulting conclusion are apt to be most untrustworthy. An important feature of any investigation is not only the discovery of the probable source but the ruling out of other possible sources. While one

may carry the salient features of each case in mind for a time, unforeseen questions sometimes arise which memory can not answer. A complete investigation is desirable for the same reason. While it is usually advantageous in a large outbreak to see a considerable number of cases within a short space of time this should be followed by a more careful inquiry at the earliest opportunity. In the preliminary investigation we seek the possible source that is common to, say, the first ten cases and simply inquire of another ten whether or not they were exposed to this source of infection. The purpose of this, as must be apparent, is solely to insure that this highly probable source shall be taken care promptly. However, anyone who has encountered the "show-me" attitude of the well owner, or the dairyman whose property, or the fond mother whose child, is accused of transmitting disease, will appreciate the comfort that' complete written records can give. It is only by a record of the negations as well as of the affirmations that one can refute the sometimes plausible, sometimes fanciful, theories advanced by interested persons.

of

informant to remember, but not infrequently it is for the deliberate purpose of misleading. The investigator can generally determine whether it is purposeful or not, provided he discovers that the statements are incorrect. It is in order to check the more important items of information that one usually asks two or three questions along the same lines and compares the answers. For example, one inquires the date of onset of a case of scarlet fever; we are told it was on Wednesday, the 12th of the month. We ask the dates of going to bed and of the physician's first visit and learn that they were on the same day, but a reference to the health officer's record shows that the case was reported on Thursday, the twentieth, and that the case was quarantined that same day. Other incidents may then be called to mind which substantiate the later date as the correct one. Other items may be confirmed or their fallacy demonstrated in much the same fashion.

To overcome the handicap resulting from the deliberate perversion of truth, one must usually adapt himself to the circumstances before him. As an instance the following experience may be related. A number of typhoid cases had suddenly appeared in a small town, all on one milk route. The source of infection, however, could not be located. Finally the health of

ficer, accompanied by a dairy inspector, visited the dairy farm for perhaps the fifth or sixth time. While the former engaged the family in conversation, the inspector wandered about the farm. The latter was an Irishman of no little wit and acuity and in his promenade he looked over the employes. Picking out the dullest-looking lout among them, he approached him as he was at work in a field and asked, "Where is the sick man?" The dull one replied that the sick man was gone; but beyond the fact that he had left a week before he could give no information. However, with the information in his possession, the inspector was able to secure a complete confession from the proprietor and the case was located in a hospital in another

state.

The tabulation and interpretation. of data suffer oftentimes by default, sometimes through a disregard of statistical methods, or through loose reasoning. Frequently the fault lies with the data collected, which is insufficient and. difficult to tabulate. This is best avoided through the use of schedules or questionnaires, the investigator having a thorough understanding of the character of the information sought. The collection, tabulation, and interpretation of data, together with a description of the forms now being prepared by the State Department of Health will be considered in a later issue.