It must be admitted that Cabot's findings are discouraging. They are not so bad as they would seem, however, at first thought. Death certificates, unfortunately, do not have the significance which they should have. Physicians are well known to be entirely too careless in their preparation and inclined to look upon them merely as legal formalities which cannot readily be avoided. It is furthermore difficult, as every doctor knows, to point to one immediate primary cause of death in every instance. Very often there is a combination of factors concerned and it is possible at practically every autopsy to find lesions not represented in any way whatever in the death certificate. It is unquestionably true that statistics of any kind must be based on information some of which we know to be inaccurate. This should not be used as an argument for discontinuing, absolutely, our search for knowledge. It is merely a reason why our clinical standards should be improved.

An exceedingly important contribution to our rather limited fund of accurate information regarding the general health of the country was the publication recently issued by the Metropolitan Life Insurance Company[2] on the mortality statistics of wage earners and their families. This covers a period of six years (1911 to 1916) and represents a study of 635,449 deaths. The cases reported came from every state in the union with the following exceptions: Mississippi, North Dakota, South Dakota, Wyoming, Colorado, Texas, Nevada, Arizona and New Mexico. Canada and many other localities outside of the "Registration Area" of the United States Census Bureau were included. The facts presented in this report are unique in that they render available for the first time a careful and detailed consideration of the diseases which may be looked upon as representative of the industrial population of the country. The various occupations shown in the order of their numerical importance were as follows:—Laborers, teamsters, drivers and chauffeurs, machinists, textile mill operatives, clerks, office assistants, etc. It covers a study of ten million policy holders and nearly fifty-four million years of life in the aggregate. The age groups studied range from one year to seventy-five in ratios not very different from those exhibited in the general population. The death rate for all persons exposed was 11.81 per 1,000 as compared with a rate of over thirteen per 1,000 (white) of the general population of the registration area during the same period of time. The death rate per 100,000 from 1911 to 1916 of some of the more important general diseases was as follows:

Typhoid fever16.8
Diphtheria and croup24.3
Scarlet fever8.6
Acute articular rheumatism6.3
Diabetes14.4
Cancer and other malignant tumors70.0
Bronchopneumonia30.2
Diarrhea and enteritis (over two years old)13.9
Cirrhosis of the liver15.0
Puerperal septicemia8.1
Accidents of all forms75.1
Ill-defined diseases10.1
Measles8.9
Influenza15.0
Tuberculosis (all forms)205.1
Tuberculosis (pulmonary)173.9
Alcoholism4.7
Diseases of the arteries, including atheroma, aneurysm, etc.17.0
Pneumonia (lobar and undefined)77.5
Intestinal obstruction5.9
Bright's disease96.8
Suicide12.2
Homicide7.0

The death rate for syphilis, locomotor ataxia and general paralysis of the insane, combined, was 14.3 per 100,000. The percentage of deaths due to diseases of the nervous system, many of which must be looked upon as probably having been associated with mental disturbances, is somewhat surprising, as shown by the following table:

Encephalitis1.0
Meningitis7.8
Locomotor ataxia1.5
Acute anterior poliomyelitis3.5
Other diseases of the spinal cord4.0
Cerebral hemorrhage (apoplexy)68.1
Softening of the brain.9
Paralysis without specified cause5.2
General paralysis of the insane4.1
Other forms of mental alienation1.4
Epilepsy3.5
Convulsions (non-puerperal).2
Chorea.2
Neuralgia and neuritis.6
Other diseases of the nervous system2.5

This shows a total rate of 104.5 per 100,000 due to diseases of the nervous system. If to this we add those dying of senility and the suicides as probably representing psychoses it would bring the total up to 123.2 per 100,000. It must be confessed, however, that such speculations mean comparatively little.

Practically the only other source of information at our disposal relative to the incidence of general diseases in the community is the tabulation of communicable diseases by Boards of Heath. The annual report of the United States Public Health Service for 1919 shows a case rate for diphtheria of 137 per 100,000 of the population based on the reports of thirty-seven states. The case rate for measles in thirty-seven states was 170. Poliomyelitis in thirty states showed a rate of 2.5 and scarlet fever a rate of 110 in thirty-seven states. The smallpox rate was sixty-eight and represented thirty-six states. The typhoid fever rate for thirty-seven states was only forty. The case rate for tuberculosis, all forms, was 346.7 in 1918. It was 274.2 in New York, 271.6 in the District of Columbia and 271.3 in New Jersey. These were the highest reported in the United States during that year. Unfortunately these statistics relate to communicable diseases only. This difficulty is due largely to the fact that comparatively few states have made attempts to keep elaborate records. The reports of Massachusetts are probably as comprehensive as any. The case rate per 100,000 of the population of all reportable diseases during the year 1920 was as follows:

Influenza938.5
Measles830.7
Pneumonia, lobar143.6
German measles12.5
Pulmonary tuberculosis173.1
Tuberculosis, other forms20.7
Diphtheria194.2
Gonorrhea186.7
Whooping cough258.3
Scarlet fever265.2
Chicken pox138.4
Mumps154.1
Syphilis77.2
Ophthalmia42.3
Typhoid fever24.2
Dysentery1.0
Epidemic cerebrospinal meningitis4.7
Malaria1.6
Pellagra.4
Smallpox.7
Trachoma2.2

The case rates for influenza and pneumonia cannot be looked upon as representative, owing to the epidemic of 1919 and 1920. During 1917 the death rate from influenza was 12.9 per 100,000 and from pneumonia 163.8. The death rate from heart diseases (organic diseases of the heart and endocarditis) in Massachusetts in 1920 was 178 per 100,000 of the population, from apoplexy 108.4, cancer and other malignant diseases 116.7, Bright's disease and nephritis 92.4, diarrhea and enteritis 52.9, violence 76.3, automobile accidents and injuries 11.9 and suicides 10.1.

It must be admitted that it is exceedingly difficult to establish a definite basis for a comparison of our statistics relating to mental disorders and those dealing with the frequency of other diseases in the community. As has been shown, our information on the latter subject, such as it is, has to do only with communicable diseases and the reported death rates. In making an analysis of the reports of mental diseases we are limited almost entirely to the institution population. It is true that these statistics are much more reliable than the others, as we are dealing with a stable population entirely under control. The cases, furthermore, are almost invariably subject to a prolonged observation and careful study. The diagnosis in almost every instance is based on elaborate mental examinations and exhaustive personal and family histories. It is, of course, true that there are innumerable cases of mental diseases outside of institutions. There were 18,268 patients at home on visit from the state hospitals alone on January 1, 1920. Those not requiring hospital treatment or custody in an institution can, however, be eliminated for the purpose of comparative studies. The fact that an analysis of death rates alone does not throw any light whatever on the frequence of psychoses for reasons already given will, I think, be conceded. For statistical purposes, at least, it may be assumed that the frequence of mental diseases as shown by a study of the hospital population is fairly representative of conditions existing in the community.