ing knowledge of epidemic diseases, such investigations, often of necessity carried out at a distance from the field, never have given, and perhaps never will give, a complete knowledge of the conditions governing the spread of epidemic diseases. First-hand knowledge of attendant conditions, derived from observations in the field, have always been necessary to give a practical solution to the problem of the control of any epidemic disease; and this is especially true in regard to epidemic poliomyelitis, which seems in so many respects to disregard the laws which are supposed to govern epidemics of contagious diseases.

MORBIDITY REPORTS.

It is of the utmost importance to ascertain the exact prevalence of the disease. To accomplish this it is absolutely essential that the disease be made reportable in all States. The transmissibility of epidemic poliomyelitis has already been sufficiently indicated to justify such a requirement on the ground of protection to the community; and as a means of obtaining accurate statistics the measure is absolutely essential. Laws to this effect have already been made in a number of States, and it is to be hoped that in the coming year all other States will follow their example.

So far the disease has been made reportable chiefly, if not solely, in States where its prevalence has already alarmed the people. Other States should not postpone their legislation until such circumstances make it imperative, but should at once enact laws to keep them forewarned and forearmed.

The importance of obtaining reports of all cases of anterior poliomyelitis may be illustrated by a few examples:

1. Our knowledge of its prevalence is at present derived largely from unofficial

reports of epidemics. These reports embrace for the most part only outbreaks of sufficient magnitude to have attracted special attention and study, failing very often to take account of scattered socalled sporadic cases. The result is a failure to give an accurate idea of the actual prevalence of the disease and, what is perhaps of greater importance, a failure to grasp the connection between seemingly isolated cases and epidemic foci. A case which appears absolutely isolated to the attending physician or even to the local health authorities may be seen by the State health officer, who has before him reports of all cases in the State, to have a definite relation to some epidemic focus.

2. By reports of all cases, the isolated as well as the epidemic, valuable inferences may be drawn as to the influence of many large factors, such as density of population, routes of travel, climatic conditions, drainage, the prevalence of insects, the prevalence of paralysis of animals; all of these being points concerning which the most careful intensive study of epidemic foci alone is apt to give erroneous impressions.

3. Prompt and accurate morbidity reports are obviously necessary as a preliminary to intensive study of cases. An edict making poliomyelitis reportable in Sweden laid the foundation for the epidemiological study of poliomyelitis, making possible the extensive studies of Wickman.

Reports from a large area of country cannot be expected to be accurate in detail. Such reports must necessarily be obtained from hundreds of different observers, each introducing an unknown coefficient of error in his own personal bias. To reduce this error, such extensive reports should be made as simple as possible, embracing only bare facts, in re

porting which the chances of error due to faulty observation, carelessness in 'expression, or unwarranted inferences are reduced to a minimum. Much will be lacking in these reports, much that is of importance in interpreting the laws of epidemic poliomyelitis; but they will at least have the advantage of being broad and, what is better, of being accurate.

INTENSIVE FIELD INVESTIGATION.

To supplement the extensive knowledge gained by collective reports, it is necessary to have other observations not less accurate, but more detailed. These observations must be made by individual intensive studies, in which thoroughness and accuracy must be the first aim, extensiveness of observation secondary. Accuracy in such studies may best be obtained by the employment of specially trained, experienced observers; uniformity by having the men engaged in such work keep in close touch with each other; extensiveness by having a large number of observers, each of them devoting as much as possible of his time to the work. In some instances the local health officer can best make these studies, especially in small localized outbreaks, having as he does the advantage of local knowledge. In most cases, however, it is better to have the studies undertaken by the State, especially studies of epidemics so large as to require more time than the local health officer can devote and studies of cases so widely scattered as to be inaccessible to one having local duties to perform. The local health officer, can, however, even when he is not the principal in the study, be an invaluable ally, being already possessed of a knowledge of local conditions which a stranger in the community would have difficulty in acquiring without his aid.

Our knowledge of the epidemiology of

poliomyelitis is based on the result of comparatively few field studies. Wickman has contributed a careful intensive study of over 1,000 cases occurring in Sweden in 1905-6, a study which is still unsurpassed in combined extent and thoroughness. The collective investigation committee of the New York Neurological Society (9) made a careful study of the epidemic of about 2,500 cases occurring in and around New York in 1997. The Massachusetts State Board of Health has been actively engaged since 1907 in the study of the disease in that State. Their report for 1909 (1), giving the distribution of cases in the State for three years and the results of the intensive study of 150 cases, is as valuable a contribution as has ever been made to the subject and serves admirably to illustrate the advantages of combining intensive personal studies with collective reports. Minnesota has made some excellent studies on similar lines (15), the results of which have not yet been published in full. Some interesting contributions have also been made from Nebraska (10, 11), and scattered reports of similar outbreaks. from various places. During the present year the collective and intensive studies. have been continued in Massachusetts and Minnesota and similar studies undertaken in Iowa. A number of other States, including Virginia, Pennsylvania, Connecticut and Kansas, and doubtless still others have undertaken at least collective studies of the disease, while in the District of Columbia a collective study has been undertaken by an organization. of the medical profession.

The information gathered from the studies in 1910 will be very valuable, but still not sufficient. Reports are wanted from every State to give a clear idea of the situation and how to control it.

METHODS OF INTENSIVE FIELD STUDY.

To take up now in detail the objects, methods and difficulties of an intensive study of epidemic poliomyelitis:

COLLECTION OF CASES.

The official morbidity reports must first be verified as to accuracy of date and diagnosis. Almost invariably, too, these reports will have to be supplemented by the addition of abortive and suspected cases. It is not even to be expected as yet that official reports will include all the abortive cases of poliomyelitis occurring in a community, although the wide discussion of the subject now taking place, calling attention to the existence of such cases, will undoubtedly result soon in their more general recognition.

Wickman (7), in reporting his exhaustive studies of epidemic poliomyelitis in Sweden in 1905-6, first pointed out clearly the occurrence of abortive forms of the infection and emphasized strongly their frequency and epidemiological importance. He distinguished several types of abortive cases.

1. With symptoms of general infection.

2. With symptoms indicative of meningitis.

3. With hyperæsthesia and pain. 4. With gastro-intestinal disturbances. Cases showing symptoms referable to the central nervous system, such as meningitis, hyperæsthesia, disturbances of reflexes, or transitory paresis, are sufficiently distinctive to make a clinical diagnosis possible. Other cases, however, can be diagnosed only by inference, from their relation to typical cases of poliomyelitis, and are almost certain to be overlooked unless this relation is known. The practicing physician is usually unaware of the relation of his cases to cases occurring in the practice of other physi

cians. Prompt reporting of all cases to the local health officer will therefore not only help the health officer, but will equally help the practitioner who, by keeping in touch with the health officer and being informed of the relation between cases, may often get a lead on an otherwise impossible diagnosis.

Caverly (12) states that, during the epidemic of poliomyelitis observed by him near Rutland, Vt., in 1894, the prevailing diseases of children were accompanied by unusual nervous symptoms; and similar observations have been made in other epidemics. It would be of great value to obtain, in each focus of epidemic poliomyelitis, careful information concerning diseases of children diagnosed as influenza, neuritis, muscular rheumatism, "summer complaint," etc. Such information can be obtained only by enlisting the hearty co-operation of practicing physicians.

Very frequently, also, abortive cases of poliomyelitis are so slight as not to have. been brought to the attention of any physician. The matter, then, of tracing out abortive cases is always one of difficulty, and there is good reason to believe that, except in very limited epidemic foci, such cases have never been traced with satisfactory thoroughness. A house to house canvas of the town seems the only way to accomplish this end satisfactorily.

After tracing up possible abortive cases of poliomyelitis there remains the even greater difficulty of deciding which of these cases may be safely considered as due to this infection. There is the danger on the one hand of too great conservatism and on the other hand of too great enthusiasm for a convenient diagnosis. On the whole, I think it may be safely asserted that the error has generally been on the side of conservatism. In order that

the epidemiologist may be able to decide which cases he shall include under the diagnosis of poliomyelitis, it is necessary that he should make a careful clinical study of the disease and that he should, if possible, be provided with a field laboratory sufficient to enable him to make examinations of blood and cerebro-spinal fluid. Examinations of this kind promise to be very helpful to the epidemiologist in the future. Especially in regard to abortive cases it is highly important that the field study be undertaken during the progress of the epidemic or very shortly thereafter, as such mild cases of illness will often have been forgotten alike by physician and family within a very few weeks after their occurrence.

It may not be out of place here to call attention to the frequency of abortive, as compared with paralytic, cases in several different localities.

Of the 1,025 cases studied by Wickman (7) in Sweden during 1905-6, 157, or a little over 15 per cent., are classed as of the abortive type. The author states, however, that this does not in his opinion represent the true proportion of such

cases.

In three circumscribed epidemic foci, offering favorable opportunities for tracing all cases, Wickman found 68 paralytic cases and 62 of the abortive type, approximately 48 per cent. of the total. Taking into consideration only those houses in each of which there occurred more than one case, Wickman states that of 404 cases occurring in 156 houses, 211, or 52 per cent., were of the abortive type.

In Massachusetts (1), in the intensive. study of 150 paralytic cases occurring in 142 houses, 49 possible abortive cases were found to have occurred in the same houses, making 26.6 per cent. of the total

cases.

In a field study in Iowa during the past

summer the writer investigated 67 houses in which there had been 74 paralytic cases and 44 possible abortive cases, making a total of 118 cases, of which 37 per cent. were possible abortive types. Taking into consideration cases occurring in the same vicinity but not in the same house with paralytic cases, I collected 83 cases which I suspected to be abortive types of poliomyelitis, as compared with 74 frank

cases.

Anderson (11), in a summary of 86 cases occurring in Polk county, Neb., in the summer of 1909, states that 40 per cent of the cases showed no paralysis.

Muller (13) gives an account of an epidemic, evidently poliomyelitis, occurring in the island of Nauru, in Oceania, in January, 1910. Within two weeks 700 of the 2,500 inhabitants of the island were attacked by an acute general infection affecting the nervous system, but of these 700 only about 50 showed paralysis after three months.

The occurrence of abortive cases of poliomyelitis is by this time well established, and while conservatism in diagnosis is to be commended, we can no longer make definite and lasting paralysis the criterion for inclusion of cases under the diagnosis of poliomyelitis. Abortive cases may be considered as probably more important than paralytic cases in the epidemiology of this disease, and no intensive study can now claim to be complete without taking such cases into consideration. These cases, in fact, are deserving of special study, both by the clinician and the epidemiologist.

LOCATION OF CASES.

The plotting of cases upon a map is a helpful and even necessary procedure. The map should be as nearly as possible accurate, and should be on a generous scale. The cases should be plotted on this

map with care as to location and with an easily comprehended graphic representation of the date as well as the location of each case. Such a map, showing at a glance the grouping of cases with regard to previous cases, as well as in relation to elevation, drainage, sewage, disposal, dusty streets, etc., often show more at a glance than could be learned from the study of many tabulations.

The map, however, is often misleading unless interpreted in the light of further observations. Epidemiological observations to be reliable must be made by personal canvass of cases. Allowance must be made for a certain amount of error in the information obtained from even the most careful personal canvass. It is 'the realization of this unavoidable error which leads those who have tried to get accurate information by this means to distrust the accuracy of compilations made from the scattering observation of many different observers

SYMPTOMATOLOGY.

In the canvass of cases of poliomyelitis it is necessary to go into the symptomatology of each cases with more care than is usually required in the epidemiological study of other infectious diseases. This is necessary because, as already stated, in many cases the diagnosis is doubtful, and clinical study is necessary to give to these cases their proper epidemiological significance. It is desirable also to utilize such an opportunity to collect data as to the symptomatology and ultimate effects of epidemic poliomyelitis.

CONTACT.

In trying to determine the source of infection in each case, while no possible factor should be overlooked, special attention should be paid to determining contact with previous cases, paralytic or

abortive. Even when there has been direct contact with a previous case in the acute stage of the disease, it is not always easy to determine this. Contact with unrecognized abortive cases is still more difficult to determine, especially in the case of children, whose playmates are often unknown to the parents. In reckoning the chances for contact account must be taken of neighbors, chance playmates, visitors and schoolmates; also attendance at schools, Sunday schools and church, public places of business or amusements, railway travel, public drinking cups, etc. Add to this the chances of indirect contact through other members of the family, visitors, servants, tradesmen, etc., and the possible avenues of contact become surprisingly numerous and complex, even for a child kept strictly at home in a small family comparatively isolated. Complicate all this with confusion of dates, failure to remember visits and visitors, and all the other vagaries of the memory, and it is readily seen that even the most careful investigator must needs be very cautious about asserting that there was no chance of contact infection in any given case.

Considering then the difficulties of tracing contact between cases, the tracing of contact is of more epidemiological value. than the failure to trace it. This is especially true as regards many of the epidemics which have been reported after very superficial observation.

On the other hand, in interpreting the finding that a certain percentage of cases have been in contact with previous cases, it is necessary to take into consideration. numerous factors, such as the probable number of persons exposed to infection and the proportion of these that develop the disease. For instance, in a small community where there had been, say, one case per hundred inhabitants, it would