FLIES AND DIARRHEAL DISEASE / SECOND YEAR'S INVESTIGATION FIVE FACTORS IN DIARRHEAL DISEASE Percentage Distribution of Infants and Percentage of Attacks of Diarrhea Distributed According to Fly Protection, Care of Home, Type of Feeding, Age, and Nationality. Based on Selected Group. FLIES AND DIARRHEAL DISEASE SECOND YEAR'S INVESTIGATION A JOINT STUDY BY THE BUREAU OF PUBLIC HEALTH AND HYGIENIE OF THE A. I. C. P. AND THE BUREAU OF CHILD HYGIENE OF THE DEPARTMENT OF HEALTH The New York Association for Improving the Condition of the Poor 105 East 22nd Street New York City Publication No. 91 THE NEW YORK ASSOCIATION FOR IMPROVING THE CONDITION OF THE POOR CORNELIUS N. BLISS, Jr., President BAILEY B. BURRITT, General Director BUREAU OF PUBLIC HEALTH AND HYGIENE OF THE DEPARTMENT OF SOCIAL WELFARE HAVEN EMERSON, M. D., Chairman LEO ARNSTEIN HOMER FOLKS J. H. HUDDLESTON, M. D. FREDERIC P. MOORE WILLIAM H. PARK, M. D. HERBERT D. PEASE, M. D. GEORGE C. WHIPPLE C-E. A. WINSLOW DONALD B. ARMSTRONG, M. D„ Director PHILIP S. PLATT, Superintendent CONTENTS „ , PAGE r oreword 4 The Problem : Method of Attack and Conclusions . 5 Flies and Diarrheal Disease 7 Plan of Study: 1. Census 8 2. Description of Protective Measures 10 Plan of Analysis of Study 12 Social Data 15 Separation of Nurses into Two Groups 17 Results: Discussion of Tables of Selected Nurses 19 Analysis within Protection Groups 25 Discussion of Tables of Combined Selected and Non-Selected Nurses 32 Modified Conclusions 33 Tables of Non-Selected Group 35 "fables of Combined Group 41 FOREWORD This study was made jointly by the Department of Health of New* York City and the Bureau of Public Health and Hygiene of the New York Association for Improving the Condition of the Poor. It is one more ex- ample of the effective co-operation of public and volun- tary activities in attacking a common social problem. Social-sanitary problems may be approached from several angles. On one hand lie the accepted theories of the day and the strongly expressed statements of the propagandists. On the other hand, highly specialized methods of statistical analysis offer the most precise and accurate measurements that can be made. Since such studies demand a very special mathematical technique they are necessarily few in number. A third source of information is illustrated by this report. Dealing with fairly large numbers under careful control this method of analysis while not pretending to extreme accuracy can furnish conclusions of very significant value. It is to such a study that we commonly look for direction and justification in our extensive social and sanitary wel- fare work. The need of subjecting opinions on such a popular question as the importance of flies in transmitting disease, to the critical test of careful investigation before effective social action can be taken, is apparent. This report is submitted in the hope that the contribution it makes to present day knowledge regarding the relation- ship between flies, dirt, feeding and diarrheal disease among infants will make possible more intelligent and efficient social efforts to reduce the amount of sickness and death from such diseases among children. Throughout the study the Bureau has received from many sources most generous and helpful advice. With- out the generous assistance of the Bureau of Child Hy- giene of the Health Department in providing the services of six of the nurses, the size of the study would have been greatly curtailed and its value correspondingly diminshed. To all who have assisted and especially to Dr. William H. Park, Director of the Health Department Labora- tories, the Bureau wishes to express its appreciation. 4 THE PROBLEM Method of Attack and Conclusions. Is the house fly the chief carrier of diarrhea to New York's babies, or has it more deadly rivals? Should some of the energy now expended in fly swatting be diverted to other details of home hygiene, or should the hue and cry after this insect pest be redoubled in volume? Do the fly-exposed infants suffer more than the infants in dirty homes, or than infants who are artificially fed? An answer to these important questions has been sought in a two-years' study in the homes of more than a thousand infants, the first year's work having been re- ported in Bulletin 79. In this, the second year's experiment, the infants were divided into two similar groups and were visited every five days by twelve nurses. The fly-exposed, "con- trol" cases received all the instruction usually given in child hygiene work but no special emphasis was laid on protection against the house fly. In the protected group, on the other hand, in addition to the instruction given the first group, great emphasis was laid on the screening of the baby; in the cradle, go-cart, on the bed and even in the arms, constant use of netting was insisted upon. Over one thousand yards of netting were distributed among the protected families. Effective fly literature, distribution of fly paper and special instruction by the supervisors were other means of reaching the families in which protective measures were taken. Accurate rec- ords of the health of the infant, the use of netting, feed- ing methods and other pertinent information were made at each visit. Analysis of some eighteen thousand rec- ords was made from the point of view of fly protection or exposure, home sanitation, feeding methods, nation- ality, age and atmospheric conditions of temperature and humidity. The conclusions indicate clearly the relative import- ance in infant welfare of flies, dirt, and artificial feeding 5 as determined by this study. Furthermore, the import- ance, in infant hygiene of each condition in itself is strongly established. In baby welfare work none can be neglected. CONCLUSIONS.* Fly Factor 1.9 Almost twice as many infants (1.9) had diarrhea among fly-exposed as among the fly-protected infants. The group of influences, associated with a dirty home, and designated as the "dirt" factor, play a similar part in diarrheal incidence among infants. Almost twice as many infants (1.8) had diarrhea in dirty homes as in clean homes. Of somewhat greater importance is the influence of artificial feeding. Almost two and a half times as many infants (2.4) were attacked by diarrhea among the arti- ficially fed as among the breast fed infants. If the milk had not been selected and instruction in feeding given in many cases, this difference would have been still greater. The influence of flies and "dirt" combined is of sim- ilar importance to that of artificial feeding. Almost two and a half times as many fly-exposed infants (2.4) in dirty homes were attacked by diarrhea as fly protected infants in clean homes. The influence of "dirt" and artificial feeding com- bined is still greater. Three and a half times as many artificially fed infants (3.5) in dirty homes were at- tacked by diarrhea as breast fed infants in clean homes. " Dirt " Factor 1.8 Artificial Feeding Factor 2.4 Flies and " Dirt " Factor 2.4 " Dirt " and Artificial Feeding Factor 3.5 *Based on work of Selected Nurses. (See pages 17-19.) 6 FLIES AND DIARRHEAL DISEASE. By Philip S. Platt, Superintendent Bureau of Public Health and Hygiene. The influence of flies in causing diarrhea among in- fants under one year of age in New York City, with special reference to the sanitary condition of the home, type of feeding, and the age of the infant, is the sub- ject of the following report. The problem is one in which loose reasoning and easy assumptions have been the rule rather than the exception. The propagandist has made much of extreme and unreliable statements. Little if any serious attempt has been made to measure the effect upon the fly prob- lem of such important influences as the sanitary con- dition of the home, type of feeding and the like. The fly has often been given the credit of the combined in- fluence of these several factors. Is the fly as important in clean homes as in dirty homes? What is the influence of flies and "dirt" combined? Such questions are funda- mental to any study of the fly. Relative sanitary values are of the utmost importance. We know of course that there are sanitary problems peculiar to one climate that are insignificant in other cli- mates. The hookworm, yellow fever, and sleeping sick- ness, for instance, are diseases only slightly if ever en- countered in the northern climates. The fly in its very nature as an insect-carrier of disease is more or less im- portant according as conditions are favorable or unfavor- able for the growth of the insect and opportunity for ac- cess to the bacteria of intestinal origin is considerable or slight. While there is every reason to believe that the importance of the fly as a carrier of disease is not the same in a well-sewered northern city as in southern cities and towns with greater prevalence of open privies and longer fly season, there have been to our knowledge no properly controlled fly experiments in the northern The Need Especially in the North 7 cities. If the health officers in the north are to deter- mine accurately what results can be accomplished with the resources at their disposal there is an ever greater need of carefully controlled measurements of such sani- tary problems as the importance of the fly. Without a sense of proportion regarding sanitary values, an indict- ment against the fly whether light or heavy is of dubious significance. It was with the intention of determining the influence of the fly under these carefully controlled northern con- ditions that the Bureau of Public Health and Hygiene devoted the summer of 1914 to an intensive study in the homes of over a thousand infants. The subject is not one with which the Bureau is unfamiliar. The previous summer (1913) was likewise devoted to a study of the relationship between flies and diarrheal disease. In this experiment, while three times as much diarrhea occurred among those of all ages in a fly-exposed, unsanitated area as in a fly-protected, cleaned-up area, no attempt was made to isolate or meas- ure the several influential factors and the results conse- quently were not accredited to the fly alone. Recogniz- ing the preliminary nature of this study and the need of substantiation by more delicate methods and more care- ful control, the present experiment was planned. While the period of observation was from June 9 to September 15, the period covered from the time the plans were laid until the preparation of the report was a full year. Purpose Previous Study I.-PLAN OF STUDY. Size A frequent criticism of social investigations is justly made upon the grounds of the smallness of the numbers involved. In order to meet this criticism, an endeavor was made to conduct this experiment on as large a scale as the possibility for thorough work permitted. Happily, the size of the experiment was further augmented by the fact that it was conducted jointly by the Department of Health and the Association for Improving the Condition 1.-Census. Health Depart- Nurses 8 of the Poor. In view of the direct importance of the study upon the problems of infant welfare in New York, the Health Department gave the services of six of the Department Milk Station nurses, who were able, how- ever, to combine their own work with that which our study involved. Six nurses employed by the Bureau, brought the number to twelve. Estimating that 100 visits a week should be made, a census of 1,200 homes containing an infant under one year of age was planned. The selection of the districts in which to make the study was given careful consideration. In order to make the study representative of more than one nationality, and get a racial balance, Jewish, Italian, and Irish dis- tricts, with uniformly high population congestion and diarrheal rates, were chosen. Two contiguous down- town districts, consisting mainly of Italian and Jewish families and bounded by Water, Henry, James and Mont- gomery Streets were selected. An uptown district on the West Side, bounded by 46th Street, 54th Street, 9th Avenue and the Hudson River, was largely of Irish na- tionality. Of course, the districts are by no means of single nationality and infants were not excluded because of other nationality than the one predominating. In these three districts the canvass for the 1,200 in- fants under one year of age was begun on June 9, and lasted until June 21. A census card had been prepared upon which full data regarding the condition of the home, of the family, and of the infant were recorded. Special emphasis was laid on the care of the child and the willingness to co-operate in special measures, such as screening of baby and food, covering of diaper can, etc., the replies being properly graded so as to indicate the degree of co-operation which might be expected. Such social data as age, nativity, time in United States, previous history of the family, sickness, deaths and pres- ent conditions of health were also recorded. The analysis of these census cards showed 1153 homes with 1165 infants under one year of age, and a total of 2130 children under 5 years distributed almost equally in the three districts. These, of course, were A. I. C. P. Nurses Selection of Districts Census Division of Census 9 divided at once into two groups as nearly identical as possible. In one group the special fly-protective meas- ures were to be enforced; in the other group the absence of such measures would enable it to serve as a "control". Great pains were taken to make the two groups entirely comparable. Many conditions were taken into considera- tion in this division; namely, the size of the family, the care of the infant, the promise of future co-operation, the possession of fly netting for the baby's protection, and the presence of sick or delicate children, of which there were 6%. After division, the districts were practically iden- tical in all these respects. These infants were then dis- tributed equally among the twelve nurses on the basis of their familiarity with the people or the language of any district. 2.-Description of Protective Measures. Screening of Windows not Attempted The group to be protected against the fly has been mentioned. Immediately the question arises as to the nature of the protection. At first thought, screening of windows occurs as the chief if not only protective meas- ure of importance. It was upon this point that the ex- perience gained in our study of the previous summer was of very practical value. For the women of the lower East Side, a window possesses an importance of the first magnitude, and the suggestion of screening a window in such a way as to curtail ever so slightly its important social function will not readily be entertained. Even sliding screens, the installation of which amounts to a very considerable sum, were found to be quite im- practicable. However, even if the screens were properly used and cared for, the influence which they would exert on the children from one to five, who are so largely in the streets and eating all kinds of street food, is probably not great. Consequently, it was decided that no attempt would be made to protect the child from 1 to 5 years of age, but that every effort should be concentrated on the infant under one year of age. The infant in its carriage or its go-cart, cradle, or bed can be effectively protected against the fly by the proper use of netting. It was upon Netting Over Baby 10 V J U«o LX < £ X Lu kJI U.^^jXILK w ■ i^' ',-' ^R ia>»«un>«i ■ VfTlAU^ K'X O X^l *y-Im z. y WM1 _ --- r T 1 ^TrM^ht . £ *j Stable ^1 ■■■■■■■|^■■ J / HKxJnified > ■ jylegj^^ ZWfeL\ ^'■1 bU w ■ 'S»4s iC I »v<.tU'iJ BlSSCBSEH । ■myi | cannEd. M||||||M Jr i " "' " VZT^ -- /-- . kT" r----*-^*^ ■* Sh^-QkS 1 (X Bl i ® I wwvIKSH H ^hc 9 ' L ATimbfeit^ j8 unseen lidef THE SINEWS OF WAR AGAINST THE FLY Specimens of Illustrated Circulars, Netting, Fly Paper, Traps and Inspection Cards used During the Summer. this that the greatest emphasis was laid. The insistent demand was that the netting must be over the baby. In order that there would be no tax upon the slim resources of these families, the Bureau provided the netting. About 1000 yards of netting were thus distributed. In addition to the distribution and use of netting, five different types of popular fly literature, effectively illustrated, were placed in these homes. Tanglefoot fly paper, generously contributed by an interested firm, was distributed each week, chiefly as a means of further fix- ing attention on the fly. Of course, the visiting of the nurses was of the greatest importance, not only because of the necessity of recording the conditions existing in the home each week, but also to give the stimulus, en- couragement and advice to keep up the interest and en- thusiasm of the mother. On the other hand, in the so-called unprotected or "control" group, the special measures with which the fly was combatted, were not emphasized. Netting, litera- ture, and fly paper were not distributed, though the nurses made an equal number of visits. In making these visits the nurses while not dwelling upon the fly, ob- served all the conditions existing in the home and gave the usual sanitary and hygienic instruction customary to the Health Department nurses. Of course, when any case of sickness was discovered, regardless of the group, every effort was taken to restore the sick one to health. Children with diarrhea were sent to the milk stations for further diagnosis and treatment. The information gathered by the nurses at each visit was recorded on an inspection card provided for that purpose. The nurses were required to record informa- tion concerning the health or the sickness of the children under five years of age, the use of netting over the baby or the go-cart, the exposure of food, the prevalence of flies in the rooms, the use of fly paper, the method of feeding, the handling of the milk, etc. It was upon these reports of the nurses that the final classification of the infants was made. Fly Literature Fly Paper Nurses' Visits Control Group Inspection Cards 11 Supervisors In addition to the visits of the twelve nurses, each of the three districts was under the constant observation of a supervisor who visited the protected homes, deliver- ing fly paper, reporting conditions as found and check- ing up the reports of the nurses. II.-PLAN OF ANALYSIS OF STUDY. Number of Cards At the close of the field work on September 15, the files contained some 18,000 observations upon some 1200 infants. Included in these were several hundred regard- ing whom there were insufficient reports, owing almost entirely to the removal of the families from the neighbor- hood. As many as 200 cases were thus dropped. Those remaining had a minimum of eight visits and all had been under observation at least two months. It was the rule, however, that whenever a family moved and was consequently lost, another family with an infant under one year of age should be found to take its place. This prevented the number from being too markedly reduced. Early in the experiment it became evident that the division of 1200 families into two similar groups, one in which the infant was protected against flies and the other in which flies were disregarded, created too arbitrary and inelastic a distinction to be successful. Within any group of such dimensions there must be allowance for individual variation. Within the protected group the best and the poorest degrees of co-operation were found, and in the "control" group there were to be found those who were protecting the infant against the fly in a thoroughly creditable manner. Our observation in the matter was later substantiated by comparison of the diar- rheal rates expressed in terms of the percentage of in- fants attacked by diarrhea, for the two groups. The pro- tected group as such had a rate of 22.5%, while the "con- trol" had a rate of 20.5%-a result which gives little en- couragement to the idea that real co-operation can be expected throughout a large group. Dropped Cases Classifi- cation 12 It was realized, of course, that the classification of each infant according to its particular degree of fly pro- tection must be done with the greatest care and accu- racy, else the results of the experiment would be of little worth. A method which allowed a maximum degree of reliability and the minimum of error was found. Upon the basis of the percentage of actual observations of net- ting over the baby, the netting efficiency and the degree of protection of the infant could be classified without de- pending upon any unstandardized personal impressions. The only ground for criticism exists in deciding what percentage of observations should determine good fly, fair fly, and poor fly protection. In view of the many conditions which in themselves made it impossible that the netting should always have been over the infant, such as during feeding, bathing, or while in its most ac- tive waking period, it is obvious that the netting could not always have been observed, even though the best protection was practised. After careful consideration 40% or greater of positive observations was decided upon as the range for good protection, between 10% and 40%, fair protection, below 10%, poor protection. In this way it is believed an accurate classification has been obtained. Similar methods were used in classifying the infants according to the sanitary condition of the home, and type of feeding. By the very nature of the groups, how- ever, the classification of the clean, fair and dirty homes cannot claim the same degree of reliability as is claimed for the fly-protection groups. In the first place, the judgment of the nurses necessarily played a large part, and such judgments do not allow of the most careful standardization. Futhermore, for various reasons a weekly observation on this point was not demanded, and the classifications are based chiefly on a general sum- mary of the home conditions, provided by the nurses to- wards the close of the experiment, amplified by the de- tailed statement on the census card, the information fre- quently given voluntarily by the nurses, and the im- pressions of the supervisors. Where the information was considered insufficient the card was placed in the Method of Classifi= cation Fly Protection Condition of Home 13 Type of Feeding unclassified group. Infants were classified as breast fed or artificially fed when 80% or more of the observations of the nurses were so recorded. Where the percentage was lower than 80%, as for instance, when 50% were breast, 30% artificial and 20% mixed, the case was classi- fied as a mixed feeding case. Age and nationality were two further sub-classifica- tions. The age of the infant is that given by the mother at the time the census was made. The nationality of father and mother, due to the extensive international marriages, is frequently not the same. The policy of determining the nationality of the off-spring by the na- tionality of the mother, which has been used by the United States Census Bureau, is here employed. After each infant had been classified according to the groups before mentioned, all the information record- ed by the nurses during the twelve weeks of visiting was transcribed in such form as to give a complete his- tory of the condition of the experiment week by week. The weekly observations regarding netting over baby, and over go-cart, the type of feeding, exposure of milk, preva- lence of flies and condition of the infant were summarized. The comparative efficiency of observation of the nurses could then be noted. From these large tabulation sheets the complete data regarding the infants were arranged in the various sub- classifications. This necessitated the preparation of 144 tables. The value of the method lay in the fact that any combination of groups could then be made with ease. Particular care was taken with the reports of sick- ness. These were carefully reviewed and correctly classi- fied by Dr. Anna von Sholly of the Health Department, under whose immediate control were the nurses during the field study. Diarrheal cases were separated from the non-diarrheal. Where second or third attacks of diar- rhea were reported as occurring in the same child during the summer, the separateness of the attacks was estab- lished as carefully as was possible. In the majority of cases two weeks or more intervened between any two attacks, with a report of "well" following the first attack. Age Nationality Tabulation by Weeks Tabulation of Subclasses Sickness 14 The recurrence of attacks of summer diarrhea, however, is a matter upon which it is felt, in the light of our pres- ent knowledge, no important interpretive conclusions should be based. TABLE A SOCIAL DATA. Number and Percentage Distribution of all Infants in Fly Study. Degree of FLY Tot. No. %of No. visits Aver. No. visits PROTECTION Inf. Inf. made per Inf. Good 416 45.2 4854 11.7 Fair 412 44.7 4407 10.7 Poor 93 10.1 1033 11.1 Total 921 100. 10294 11.2 Sanitary Total No. % of CONDITION OF HOME Inf. Inf. Clean 204 22.2 Fair 220 23.9 Dirty 131 14.2 Unclassified 366 39.7 Total 921 100. TYPE OF FEEDING Total No. Inf. % of Inf. Breast 636 69.0 Mixed 182 19.8 Artificial 103 11.2 - ' - - Total 921 100. AGE Total No. % of in Months Inf. Inf. 0-3 Months 279 30.3 4-6 " 232 25.2 7-9 " 240 26.0 10-12 " 170 18.5 Total 921 100. NATIONALITY Total No. % of (According to mother's) Inf. Inf. American 125 13.6 Austrian 19 2.6 English 9 1. German 25 2.7 Greek 10 1.1 Italian 283 30.8 Irish 100 10.9 Jew 308 33.4 Polish 15 1.6 Eight Other Nationalities.... 27 2.9 Total 921 100. 15 When analysis of the corrected flies was made, the number of infants comprising the study stood at 921. Table A shows the number and percentage of distribu- tion of these infants according to degree of fly protection, sanitary condition of home, type of feeding, age and na- tionality. It will be noted that about 45% enjoyed good fly protection, 45% fair (medium) and 10% poor fly pro- tection. The latter per cent, is smaller than would be desired but the fact that it is smaller shows in an inter- esting way that even though for experimental purposes one might desire it otherwise, only 10% (and of these only a little over one-half came from the original "con- trol" group) of such a population will totally neglect to protect the infant from flies. Again, in regard to the sanitary condition of the home, we have two-thirds fewer dirty homes than either clean or fair (medium) homes. Unfortunately, a large group had to be left unclassified, owing to conflicting or insufficient data on the question. However, this unclassified group does not contain a dis- proportionate number of any sub-group, as artificially fed, or poorly-protected infants. The marked preponderance of breast fed infants con- forms to the usual findings among such classes. The per- centage distribution of the three types of feeding remains very constant in all sub-groups except that of age, in which of course the percentage of artificially fed infants increases with age. When age is considered, we find a gradual decrease in number as the age increases; to wit, 30% in the 0-3 months group, 25% in the 4-6, 26% in the 7-9, and 19% in the 10-12. On account of the large extent of international mar- riages, it became necessary as before stated to classify the infants according to the mother's nationality. Some liberty has been taken, perhaps, in combining infants of Jewish, Russian-Jewish, and Russian mothers. As com- bined, however, they represent the largest group, 33%. Next comes the Italian 31%, the American and Irish fol- low with 14% and 11% respectively. Sixteen nationali- ties in all are represented. Fly Protection Condition of Home Type of Feeding Age Nationality 16 The Separation of the Nurses into Two Groups. Unfortunately, in an experiment of this nature, the quality of the work performed by a number of nurses is not likely to be of exactly the same standard. Close association with and careful supervision of the nurses necessarily gave the supervisors definite impressions in regard to the character and quality of their work. Feel- ing that the standard of accuracy of the study demanded a careful consideration of the possible unreliability of the work of some of the nurses, they were eventually divided by the supervisors into two groups designated as Selected Nurses and Non-selected Nurses, and thereafter kept entirely separate. Each group consisted of three Bureau and three Health Department nurses. The sep- aration was made at the close of the experiment, before any analysis of results was begun, and was based on the following observations of the supervisors: 1. The non-selected nurses were less interest- ed in the experiment than the selected nurses. 2. They were less willing and conscientious in following instructions. 3. Throughout the field work it was observed that the information on their inspection cards was considerably less complete, having often to be re- turned because of the omission of important informa- tion. 4. When these nurses took their vacations, con- siderably more time was lost in putting substitutes in their places. When a mathematical comparison of the actual work of the two groups of nurses was made, the results fully substantiated the general impression of the nurses. The non-selected nurses made, 1. Fourteen per cent, or 1.7 fewer visits to each infant than the selected nurses. 2. Thirteen per cent, or 590 fewer observations in regard to the prevalence of flies. Upon this evidence it must be admitted either that the selected nurses were overzealous or that the work of the non-selected nurses was materially less reliable. Reasons for Separation of Nurses Reasons Substan= tiated a. By less thorough work 17 b. By dis= cordant results Having reached by direct observation and mathemat- ical measurement an opinion regarding the relative relia- bility of the work of these nurses, comparison of the diar- rheal rates of the selected and non-selected nurses re- vealed the following facts: 1. The difference in the percentage of infants at- tacked by diarrhea and the percentage of attacks of diarrhea, as found among the selected and non-select- ed nurses, is considerable, the rates for the non-select- ed being lower in every instance. When the entire non-selected group is considered the differences in rates for infants attacked by diarrhea and for attacks of diarrhea are 9 and 10 respectively, or 36% and 33% lower than those of the selected group. 2. The diarrheal rates for the three degrees of fly protection among the non-selected nurses are in- sistent and unintelligible in the light of the selected nurses' rates. The most striking variation from what might be expected is ill the poorly protected group where-as may be seen from tables on pages 20 and 35-the differences in the diarrheal rates are 18 and 29, or 44% and 54% less than the selected group. The most detailed study within this group on the grounds of feeding, age, nationality, sanitation of home, and fly exposure, fails to explain the extreme variation. In view of the lower rates of the non-selected nurses, the question arises whether the districts visited by these nurses were healthier. Such a situation is conceivable, although the probability of such being the case is not great. There is no a priori reason for so thinking. The three districts in which the experiment was conducted were chosen for their high states of congestion and high diarrheal mortalities, and the twelve nurses were distrib- uted according to their familiarity with the language or the people of the district. On the other hand, judging by such criteria as is at hand, the evidence is that the infants under the care of the non-selected nurses should be less rather than more healthy. For example, more milk was reported as ex- Districts Similar Conditions Less Favorable for Non- Selected 18 posed (3%), while "few flies" were reported less frequent- ly (18%), and "many flies" more frequently (5%). Net- ting was found in use 7% less often and the percentage of infants in the poor protection groups was somewhat greater, 12% among the non-selected group as compared with 8% among the selected. When, moreover, the diarrheal rates for infants of the same nationality, often living practically side by side, under the care of the two groups of nurses, are compared, a marked variation is found. The rates of the four largest nationality groups composing 91% and 85% of the infants of the selected and non-selected groups, respectively, follow: Differences in Nationality Rates Selected Non-selected American 31. 11. Jew, Russian and Russian Jew Irish 33. 25. 19. 13. Italian 15. 15. Rates for above 4 nationalities 25. 15. These discrepancies further substantiate the opin- ion regarding the non-selected nurses, previously deter- mined by the quality and quantity of their work and their interest in the experiment, and would seem to justify the exclusion of their work from equal consider- ation with the selected nurses' work. In determining, however, what emphasis should be given the work of these nurses, the Bureau decided that while the work of the selected nurses warranted separate interpretation and special emphasis, the results obtained by the selected, the non-selected, and the combined nurses should be fully presented and that modified conclusions be drawn if necessary from the combined results. Primary Conclusions on Selected Nurses Modified Conclusions on All Nurses RESULTS. The results of the study based upon the work of the selected nurses must be expressed primarily in terms of the percentage of infants attacked by diarrhea. By com- paring- the percentages of two contrasted groups the in- fluences in question may be measured. Discussion of Tables of Selected Nurses. Results of Selected Nurses 19 TABLE I.-SELECTED NURSES. Good Protection Fair Protection Poor Protection Total No. Infants. 254 185 41 % Distri. Infants. 53. 39. 8. Inf. Attack, by Diar. No. % 57 22. 45 24. 17 41. Attacks of Diar. No. % 66 26. 56 30. 22 54. Total Fly factor 480 100. 119 25. .. 1.9 144 30. FLY PROTECTION. By reference to Table I, Selected Nurses, it will be noted that 22% of the infants in the good protection group had diarrhea, and 41% of the infants in the poor protection group. The method of measuring the influence of flies is that of dividing 41% by 22%, whereby we have a factor (1.9) telling how much greater was the diarrheal rate in the one condition than in the other. It may be said then that almost twice (1.9) as many infants had diarrhea among the fly-exposed group as among the fly- protected group. It will be natural to inquire at this point whether the diarrheal rates which are given for the good, fair or poor fly protection groups are really exclusively attributable to the protection, or whether other influences such as dirty homes, or artificial feeding may not be largely re- sponsible for the high rate found, for instance, in the poor protection group. This question was examined with care, and the tests applied indicate that while occasional- ly a slight correction may be necessary, as a rule, es- pecially in the measurements of the main factors, other conditions do not influence the results sufficiently to af- fect the conclusions. It is obvious that the size of the experiment does not permit the use of highly specialized statistical methods by which every contributing influence can be accurately measured, but by the available methods, namely, the elimination of any condition, such as artifi- cial feeding, from the good protection and the poor pro- tection groups, assurance may be had that the factor is not appreciably altered. While it is impracticable to discount the age influence in every case, it appears from careful scrutiny of the sub-groups that the age factor is not of such importance as materially to affect any con- Fly Factor 1.9 Importance^ of Contribu- ting Influences 20 1 A fl fl RshWQ Were p^aCe^ ""d" oh- WAR is on! iuuu "u " servation last summer by tbe * Bureau of Public Health and Hygiene, A. I. C. P. __ __ __ and the Department of Health BABIES vs IN THE DIRTY HOMES Flies and Dirt Twice as many babies had diarrhea as in the clean homes. AMONG THE UNSCREENED BABIES WILL WIN? Twice as many babies had diarrhea as among the screened babies. t 2957 BABIES DIED OF DIARRHEA IN NEW YORK CITY LAST YEAR WHAT HAPPENED TO 1000 BABIES Dirt and flies played their part in this slaughter r\ A WELL-USED BROOM AND MOP V\ AND A YARD OF NETTING MAY SAVE YOUR BABY'S LIFE. :: :: P»rmi«ioa to reproduce MTtOMU Um (n» by E. H. Bober, author of "SwU fW* , A7&ri&D*tf> ■ F YOU LOVE YOUR BABY | A WORD TO THE WISE i 1. Keep your home clean Filth Fattens Flies Flies Kill Babies "2. Keep your food covered _ 3- Cover your garbage and make ।| B fid your neighbor do the same. Il _ ' - Cover the Food zk, « . 4- K"p ^er th. uiy. Screen the Baby 5 E't"o{°°,1,1'*,f''"1"vew'lt"1°n Clean the House •6. Buy no food from .tore, where Starve the Fly you see many Hies. JE I L, i i J\ Save the Baby 7. Report fly-breeding stables to the Health Department. F L ! E 5 , >1 Come to your house straight from the manure Fillip A pile, the water closet or the spittoon. They fWh walk on your bread, wash their dirty feet in // the baby's milk and wipe them in the sugar bowl. I IldCf* A CONTRIBUTION TO FLY CAMPAIGN LITERATURE The Results of the Study Popularized for General Distribution elusions. It should be considered also that the summer had no intensely hot periods and that there was some supervision of the food. When the question of the percentage of attacks of diarrhea is considered in contradistinction to the per- centage of infants attacked, the percentage of attacks is found to be 1.8 times greater in fly-exposed than in the fly-protected group. TABLE II.-SELECTED NURSES. Clean Fair Dirty Unclassified .. CONDITION OF HOME. Total No. % Distri. Inf. Attack, by Diar. Infants. Infants. No. % . . 78 16. 14 18. 99 21. 26 26. . . 72 15. 23 32. . . 231 48. 56 24. Attacks of Diar. No. % 16 21. 34 34. 27 38. 67 29. Total.... "Dirt" factor.. .. 480 100. 119 25. 1.8 144 30. Table II, Selected Nurses, gives the data regarding the sanitary condition of the home, determined as pre- viously described. Here it will be noted that the clean and the dirty homes are in about equal number, but the percentage of infants attacked by diarrhea is markedly different, viz. 18% for the clean and 32% for the dirty homes, or 1.8 times as many "diarrheal" infants in the dirty homes as in the clean homes. The infants in the fair (medium) homes and the large unclassified group occupy an intermediate position. The significance of the so-called "dirt" factor should be carefully considered. Of course, the classification of the homes according to clean- liness or dirtiness is based on the objective appearance of the home. Objectively, a dirty home is one in which the visitor finds the remains of a meal on the table or in the sink, dirty clothes lying about, abundant evidence of dust and dirt, and the general appearance of slovenly housekeeping. Subjectively, a dirty home probably means that the mother is of a careless, easy-going dis- position, lacking in pride and satisfied with the minimum standards of decency. Associated with these character- istics, a lack of systematic care of the infant, irregular "Dirt" Factor 1.8 Significance of "Dirt" 21 and improper feeding and other unfavorable influences on the infant are likely to be found. With this in mind, and not before, the significance of the "dirt" factor may be grasped. Probably the dirt itself is of only moderate importance; it is used here only as a symbol of a com- bination of inferior qualities on the part of the mother, and unfavorable conditions in the home. Obviously, the reverse of these characteristics, objective and subjective, are to be expected in the clean homes. While the artificially fed infants may be eliminated from the clean homes and the dirty homes without alter- ing the "dirt" factor of 1.8, the same is not strictly true of the influence of poor protection. This, however, is largely owing to the absence of infants in the clean homes of the poor protection group. Had this group of clean homes a normal size and a consistent diarrheal rate, the "dirt" factor would probably be unaffected. It may be said, however, that almost twice (1.8) as many infants had diarrhea in the dirty homes as in the clean homes. From this it appears that flies and so-called "dirt" are equally important in favoring diarrhea among infants. When the percentage of attacks of diarrhea are con- sidered, the percentage of diarrheal attacks remains 1.8 times greater in the dirty homes than in the clean. Importance of Contribu- ting Influences TABLE III.-SELECTED NURSES. TYPE OF FEEDING. Breast Fed Mixed Fed Artificially Fed. Total No. Infants. 333 101 46 % Distri. Infants. 69. 21. 10. Inf. Attack, by Diar. Attacks of Diar. No. 70 26 23 % 21. 26. 50. No. 85 32 27 % 26. 32. 59. Total 480 Artificial feeding factor 100. 119 25. .. 2.4 144 30. In connection with this table the association of a high percentage of diarrheal infants with artificial feed- ing will first be noted., the percentage being 50% as con- trasted with 21% for the breast fed. This indicates that almost two and a half times (2.4) as many infants had diarrhea among the artificially fed as among the breast Artificial Feeding Factor 2.4 22 fed. It also indicates that artificial feeding is more im- portant than either flies or dirt. The artificial feeding factor, based on the percentage of attacks of diarrhea will be seen to be 2.3. It is because of this very considerable importance of artificial feeding that it is essential to determine to what degree this influence figures in causing the high rates attributed to poor protection or dirty homes. In addition to the discussion of this question on page 20, there is another method by which the influence of artificial feed- ing in determining the other factors may be shown to be necessarily small, so small in fact that it cannot invali- date the results. It will be seen that the artificially fed group is small in number, only 10% of the whole, while the breast fed group constitutes 69% of all the infants. Consequently, if the same proportion existed in all the sub-groups, the influence of artificial feeding, with its higher diarrheal rate could never be large, and it is a fact that the distribution of the artificially fed rarely ex- ceeds 10%, except in the older age groups. Contribu- ting Influence of Artificial Feeding TABLE IV.-SELECTED NURSES. AGE. Total No. % Distri. Inf. Atta ck. by Diar. Attacks o f Diar. Infants. Infants. No. % No. % 0-3 Months. . 151 32. 29 19. 34 23. 4-6 . 120 25. 30 25. 37 31. 7-9 . 118 24. 32 27. 41 35. 10-12 . 91 19. 28 31. 32 35. Total . 480 100. 119 25. 144 30. In considering the distribution and diarrheal rates of the infants according to age, an interesting condition is observed. The percentage of infants 0-3 months of age is the largest of the four age groups while the percentage of infants having diarrhea is the lowest. Each succeeding age group has a smaller percentage of infants, and a higher diarrheal rate. The increased influence of artificial feeding with accompanying decrease of breast feeding par- tially explains this. For instance, although not shown in any of the tables, the percentage of artificial feeding for the four age periods is 3, 7, 14 and 18 per cent. re- Influence of Age 23 spectively and that of breast feeding 80, 73, 64 and 54 per cent, respectively. With this exception, the distribution of the age groups among other groups may be said to be such as not to give undue weight to the age factor. TABLE V.-SELECTED NURSES. (Grouped according to mother's.) NATIONALITY. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. Attacks of Diar. No. % No. % American . 13 3. 4 31. 4 31. Austrian . 14 3. 3 21. 5 36. Irish . 16 3. 3 19. 3 19. Italian . 182 38. 27 15. 31 17. Jew . 228 48. 75 33. 93 41. Polish 10 2. 6 60. 7 70. 4 Other Nation. 17 3. 1 6. 1 6. Total 480 100. 119 25. 144 30. Influence of Nationality Of the six chief nationalities, the smallest group had the highest percentage of children attacked by diarrhea, viz. the Polish with 60%. The next highest rate, 33%, is that of the largest group, viz. the Jewish which con- stitutes 48% of the total number in the experiment. As has been explained elsewhere, the term Jewish has been used somewhat loosely here, as constituting those whose mothers claimed Jewish, Russian, and Russian-Jewish na- tionality. The Americans come next in diarrheal rate (31%), although but a small part (3%) of the total num- ber. The Austrian and Irish are next in order, with the Italians, the second largest group (38%) possessing the lowest diarrheal rate (15%). It is felt that great empha- sis should not be placed on these nationality rates, with the exception of the Italian and the Jewish. The former are predominatingly clean and well-protected, while the latter, though no less favorably placed as regards feeding are more largely represented in the dirty and poorly-pro- tected groups. Here the figures are sufficiently large to say with some degree of accuracy that the Jewish infants are on the whole less favorably situated than the Italian. In the Jewish group 2.2 times as many infants had diar- rhea as in the Italian group. 24 ANALYSIS WITHIN PROTECTION GROUPS. Fly protection, condition of home, type of feeding, age and nationality have now been discussed singly. Discussion with reference to the other factors under con- sideration is possible, however, owing to the method of analysis which was adopted. For example, one may look within the well-protected group and see what con- ditions of home exist. TABLE VI.-SELECTED NURSES. FLY PROTECTION. ACCORDING TO CONDITION OF HOME. Good Protection. Attacks of Diar. Total No. % Distri. Inf. Attack, by Diar. Infants. Infants. No. % No. % Clean .. 55 22. 10 18. 10 18. Fair . . 55 22. 13 24. 17 31. Dirty .. 28 11. 7 25. 7 25. Unclassified .. .. 116 45. 27 23. 32 28. Total.. .. "Dirt" factor. . . . 254 100. 57 22. . 1.4 66 26. Fair Protection. Clean . . 22 12. 4 18. 6 27. Fair . . 39 21. 11 28. 14 36. Dirty . . 26 14. 8 31. 9 35. Unclassified .. .. 98 53. 22 22. 27 28. Total.... "Dirt" factor.. . . 185 100. 45 24. 1.7 56 30. Poor Protection. Clean 1 2. - - - - Fair 5 12. 2 40. 3 60. Dirty .. 18 44. 8 44. 11 61. Unclassified .. . . 17 42. 7 41. 8 47. Total.... "Dirt" factor.. .. 41 100. 17 41. 22 54. Good, Fair and Poor Protection. Clean . . 78 16. 14 18. 16 21. Fair . . 99 21. 26 26. 34 34. Dirty . . 72 15. 23 32. 27 38. Unclassified .. .. 231 48. 56 24. 67 29. Total.. .. "Dirt" factor. .. 480 100. 119 25. . . 1.8 144 30. Table VI, Selected Nurses shows the distribution and the two diarrheal rates for each type of home under the three degrees of fly protection. It will be noted that the percentage of infants in the clean homes is large under Fly Protection and Condition of Home 25 good protection and small under poor protection, while the reverse is true for percentage of infants in dirty homes. At the same time, the percentage of infants at- tacked by diarrhea within each protection group shows ascending rates according as the home conditions be- come unfavorable. Also, a marked increase of rates for similar conditions of home is observed when the various degrees of protections are compared. There is a marked consistency in the increasing rates among these twelve classifications with the exception of the poor protection, clean homes group, which is so small as to be negligi- ble. The consistency of the increasing rates for the at- tacks of diarrhea is scarcely less striking. The fly factor and the "dirt" factor, which have been discussed singly, may here be studied together. For instance, the diarrheal rate (41%) for the poor-protec- tion-dirty-homes group may be compared with the rate (18%) for the good-protection-clean-homes group. This gives a factor for flies and "dirt" combined of 2.4, a measurement which is the same as that of artificial feed- ing. Unfortunately, the plan to measure the relative im- portance of the insanitary condition of the home under good, fair, and poor fly protection as well as to obtain fly factors for clean, fair, dirty and unclassified homes, had to be given up owing to the smallness of many of the groups involved. While unwilling to give to such meas- urements the definiteness of figures, it may be stated that the work of the selected nurses shows a consistently increasing importance of the conditions generally ac- cepted as unfavorable in the diarrheal problem. For in- stance, the fly factor is obviously greater in the dirty homes than in the clean; and the "dirt" factor in the poor protection group, though it cannot be determined, gives indication of being greater than the "dirt" factor in fair or clean homes. Flies and "Dirt" Factor 2.4 26 TABLE VII.-SELECTED NURSES. FLY PROTECTION. ACCORDING TO TYPE OF FEEDING. Good Protection. Attacks of Diar. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. No. % No. % Breast Fed . 180 71. 36 20. 43 24. Mixed Fed 51 20. 11 22. 11 22. Artificially Fed. 23 9. 10 43. 12 52. Total Feeding factor. 254 100. 57 22. . 2.2 66 26. Fair Protection. Breast Fed 131 71. 26 20. 31 24. Mixed Fed 34 18. 9 27. 13 38. Artificially Fed. 20 11. 10 50. 12 60. Total Feeding factor. . 185 100. 45 24. . 2.5 56 30. Poor Protection. Breast Fed . 22 54. 8 36. 11 50. Mixed Fed 16 39. 6 38. 8 50. Artificially Fed 3 7. 3 100. 3 100. Total Feeding factor. . 41 100. 17 41. . 2.8 22 54. Good, Fair and Poor Protection. Breast Fed.. .. . 333 69. 70 21. 85 26. Mixed Fed. .. . . 101 21. 26 26. 32 32. Artificially Fed . 46 10. 23 50. 27 59. Total . 430 100. 119 25. 144 30. Feeding factor. .. 2.4 Table VII, Selected Nurses presents in similar fash- ion the data concerning fly protection and type of feed- ing combined. The figures are entirely consistent with the conclusions drawn from the data as studied separate- ly. It is unfortunate that the poor protection, artificial feeding group is so small, otherwise an interesting meas- urement of the importance of flies and artificial feeding could have been made by comparing the diarrheal rate with that of the good protection, breast fed group. The smallest group used in determining a factor has been in the group of artificially fed infants in dirty homes where 9 infants with a diarrheal rate of 56% have been compared with 55 breast fed infants in clean homes, with a diarrheal rate of 16%, giving a factor of 3.5. A Fly Protection and Type of Feeding Dirty Homes and Artificial Feeding Factor 27 measurement based upon so few cases is not of course of great value, particularly when the very considerable influence of poor fly protection is not discounted. If the artificially fed infants in dirty homes under poor pro- tection are eliminated the factor is reduced to 2.7. The artificial feeding factors show a gradual increase as the three degrees of protection are considered; viz. 2.2 for good, 2.5 for fair, and 2.8 for poor protection, with an average for the three degrees of protection of 2.4. TABLE VIII.-SELECTED NURSES. FLY PROTECTION. ACCORDING TO AGE. Good Protection. Total No. % Distri. Inf. Attack, by Diar. Attacks of Diar. Infants. Infants. No. % No. % 0-3 Months. . 91 36. 15 17. 17 19. 4-6 . 60 24. 13 22. 16 27. 7-9 . 64 25. 16 25. 20 31. 10-12 . 39 15. 13 33. 13 33. Total . 254 100. 57 22. 66 26. Fair Protection. 0-3 Months. . 54 29. 12 22. 15 28. 4-6 a . 49 27. 13 27. 17 35. 7-9 Ci . 39 21. 9 23. 11 28. 10-12 a . 43 23. 11 26. 13 30. T otal . 185 100. 45 24. 56 30. Poor Protection. 0-3 Months. 6 15. 2 33. 2 33. 4-6 . 11 27. 4 36. 4 37. 7-9 . 15 36. 7 47. 10 67. 10-12 " 9 22. 4 44. 6 67. - - - - Total . 41 100. 17 41. 22 54. Good, Fair and Poor Protection. 0-3 Months. .. 151 32. 29 19. 34 23. 4-6 . . 120 25. 30 25. 37 31. 7-9 .. 118 24. 32 27. 41 35. 10-12 .. 91 19. 28 31. 32 35. - - - - Total..... .. 480 100. 119 25. 144 30. 28 In Table VIII, Selected Nurses, showing fly protec- tion according to age, the general tendency for the diar- rheal rate to increase as the age increases and the fly pro- tection decreases is clearly seen, although the increase is by no means entirely consistent. The 0-3 group has the lowest diarrheal rate in every instance, and the 10-12 group is consistently high. Here, as described on page 23, the decrease in breast feeding with its low rate, and the increase in artificial feeding with its high rate, explains to some extent the higher rates in the older age groups. The question of age influence is, however, especially subtle. Before trustworthy conclusions can be drawn many influences should be carefully considered. The nature of the experiment, however, does not permit of such detailed analysis. Fly Protection and Age TABLE IX.-SELECTED NURSES. FLY PROTECTION. Good Protection. Attacks of Diar. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. No. % No. % American 8 3. 3 38. 3 38. Austrian 7 3. - - - - Irish 6 2. 3 50. 3 50. Italian 132 52. 19 14. 21 16. Jew 83 33. 27 33. 33 40. Polish 6 2. 5 83. 6 100. 4 Other Nation. 12 5. - - Total 264 100. 57 22. 66 26. ACCORDING TO NATIONALITY. Fair Protection. American 5 3. 1 20. 1 20. Austrian 6 3. 2 33. 3 50. Irish 10 6. - - - - Italian 46 25. 6 13. 8 17. Jew 111 60. 35 32. 43 39. Polish 3 2. - - - - 4 Other Nation. 4 2. 1 25. 1 25. Total 185 100. 45 24. 56 30. Poor Protection. American - - - - - - Austrian 1 2. 1 100. 2 200. Irish - - - - - - Italian 4 10. 2 50. 2 50 Jew 34 84. 13 38. 17 50. Polish 1 2. 1 100. 1 100. 4 Other Nation. 1 2. - - - - ■ - ' -- ■ ■ Total 41 100. 17 41. 22 54. 29 Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. Attacks of Diar. No. % No. % American 13 3. 4 31. 4 31. Austrian 14 3. 3 21. 5 36. Irish 16 3. 3 19. 3 19. Italian . 182 38. 27 15. 31 17. Jew 228 48. 75 33. 93 41. Polish 10 2. 6 60. 7 70. 4 Other Nation. 17 3. 1 6. 1 6. Total 480 100. 119 25. 144 30. Good, Fair and Poor Protection. Fly Protection and Nationality Owing to the smallness of the groups, the method of classification and the many complicating factors, conclu- sions of much value cannot be drawn from the nationality tables. In Table IX, Selected Nurses, one reason for the higher diarrheal rates of the Jewish may be seen. Fif- ten per cent, of the Jewish had poor fly protection and only 2% of the Italians. Another reason, not shown in the table, is the greater number of Jewish in dirty homes; over twice as many were in the dirty homes as was the case with the Italians. Table X, Selected Nurses, shows the average weekly mean temperatures and relative humidities, the deaths from diarrhea of infants in Manhattan, the number of new attacks of diarrhea occurring each week among the four age groups of the study, and the number of new attacks plus the continued attacks which existed each week. While there seems to be a certain association be- tween the two weeks of greatest number of diarrheal deaths and the two weeks of highest temperature and humidity, the week of highest new and current attacks is not associated with a week of particular oppressiveness. The summer was not one of exceptional hot spells, and the occasional hot days do not influence markedly the weekly averages. Three weeks reached an average of 77° Fahr., with humidities of 76, 73 and 67%. September 6- 12 was the coldest week, with 62° Fahr, and 57% hu- midity. The chart opposite page 31 shows the daily mean temperature and humidity, and gives consequently a graphic picture of the exact conditions met with during the summer. Daily Temper- atures and Humidities 30 CLIMATIC CONDITIONS AND DIARRHEAL ATTACKS June 22-27 June-July July July July July-Aug. Aug. Aug. Aug. Aug. Aug.-Sep. 30-5 Sep. 6-12 Total 28-4 5-11 12-18 19-25 26-1 2-8 9-15 16-22 23-29 Age 0-3 Months.. .. 3 2 7 2 4 3 3 5 3 2 0 34 4-6 " .. .. 0 13 4 5 6 5 3 6 3 1 37 7-9 0 2 2 2 6 8 7 6 5 1 2 41 10-12 " . . . . 2 12 7 4 4 3 4 3 0 2 32 Total 5 6 14 Total Current Attacks* 15 19 21 18 18 17 of Diarrhea by Weeks and by Age Groups. 6 5 144 Age o-3 Months.... 3 4 10 8 6 6 5 10 7 4 1 64 4-6 - 14 5 6 11 8 5 8 3 2 53 7-9 - 2 3 3 7 9 10 8 8 3 3 56 10-12 " .. .. 2 3 3 7 4 5 4 5 5 1 2 41 Total 5 10 20 23 23 31 27 28 28 11 *Obtained by adding the new attacks to the attacks continued from the proceeding week. Temperatures and Relative Humidities by Weeks. 8 214 Aver. Daily Mean Temp, Deg. Fahr. 73 66 69 77 73 69 73 77 77 70 75 62 Aver. Daily Mean Rei. Hum. in % 75 78 82 76 62 67 73 69 67 70 62 57 Deaths from Diarrhea of Infants under one year of age in Manhattan during summer of 1914. 17 20 24 45 51 73 77 86 98 83 77 66 TABLE X.-SELECTED NURSES. Attacks of Diarrhea by Weeks and by Age Groups. 31 Discussion of Diarrheal Rates of Combined Selected and Non-selected Groups. Fly Protection It will be noted from Table I, Selected and Non- selected Nurses, on page 41, that the fly factor for the combined group, obtained by dividing 31% by 22%, is 1.4, in contrast to that of the selected group alone, which is 1.9. This is the widest variation which occurs between the selected and the combined groups, and is owing to the low diarrheal rate (23%) for the non-selected, poor pro- tection group which in turn may be accounted for by the unexpectedly low rates that obtain in the poorly protected, dirty homes and the poorly protected, artifi- cially fed groups of the non-selected nurses. Among the non-selected nurses for instance, none of the 5 in- fants in the poorly protected, artificially fed group were recorded as having diarrhea. Viewed in the light of the diarrheal rates found elsewhere in the report it is hard- ly possible to consider such rates as representative, and as stated on page 19, the factor of the selected nurses is accepted as more reliable. The combined group shows a "dirt" factor of 1.9, practically the same as that for the selected group (1.8), although the actual rates for the clean, fair, dirty, and unclassified groups are in every instance considerably lower. It is interesting to note here that the lowest rates that occur anywhere in the experiment in the clean homes group. The chief reason for this appears to be the negligible influence of poor protection on the group. Among the 204 clean homes only 4 fall in the poor pro- tection group. Artificial feeding enters into both the clean and dirty groups in about the same degree, and consequently does not influence the factor appreciably. The artificial feeding factors for the selected and the combined groups are the same, namely, 2.4, in spite of a marked variation in the actual diarrheal rates. The influence of poor protection and dirty homes in deter- mining the artificial feeding factor is of no practical im- portance. Condition of Home Type of Feeding 32 The same tendency for increasing' rates with ad- vancing age is noticed, although much less markedly, in the combined group. While the temptation to subject these figures to further analysis and to determine the underlying causes is great, it is felt that such analysis would be of doubtful value. In the four nationality groups of sufficient sizes for comparative purposes, the American, Irish and Italian infants seem to have a marked advantage over the Jew- ish infants, in regard to the prevalence of diarrhea. This may be accounted for in part by the fact that while only 2% of the Jewish were in clean homes, between 23% and 30% of the other three nationalities were similarly housed. Conversely, a larger percentage of the Jewish were in dirty homes. The flies and "dirt" factor obtained by comparing the diarrheal rates of poorly protected infants in dirty homes with that of the well-protected infants in clean homes, is lowered from 2.4 to 2.1 for the combined group of nurses. The artificial feeding and "dirt" factor, obtained by comparing the diarrheal rates of artificially fed infants in dirty homes, with that of the breast fed infants in clean homes, remains unchanged at 3.5. The following table gives the five main factors as ob- tained by the selected nurses and the combined nurses: Age Nationality Flies and " Dirt " Artificial Feeding and " Dirt' Fly & Art.Feed Fly "Dirt" Art. Feed "Dirt" "Dirt" Factor Factor Factor Factor Factor Selected Nurses. . 1.9 1.8 2.4 2.4 3.5 All Nurses . 1.4 1.9 2.4 2.1 3.5 MODIFIED CONCLUSIONS. The Conclusions are based on the work of all the Nurses. While acknowledging that the reliability and accu- racy of the work of the non-selected nurses is considera- bly less than that of the selected nurses, the Bureau after due consideration decided that sufficient value was at- 33 tachecl to the work of the non-selected nurses to justify its inclusion with that of the selected nurses, especially in view of the fact that by so doing the conclusions drawn from the combined work would be based upon twice the number of infants. The modified conclusions of the study based upon the 921 infants under observation from June 9 to Septem- ber 15 differ from those based upon the 480 infants under the selected nurses in only one important particular, namely, the relative importance of flies as compared with that of "dirt". While the position of the Bureau as stated on page 19 accepts the factor (1.9) of the selected nurses as more reliable, and attributes approximately equal importance to flies and "dirt", the fly factor of the combined nurses (1.4) places the relative importance of flies somewhat below that of "dirt". Almost one and a half times as many infants (1.4) had diarrhea among fly-exposed as among fly-protected infants. The group of influences associated with a dirty home and designated as the "dirt" factor, play a more im- portant part in diarrheal incidence among infants. Al- most twice as many infants (1.9) had diarrhea in dirty homes as in clean homes. Of somewhat greater importance is the influence of artificial feeding. Almost two and a half times as many infants (2.4) were attacked by diarrhea among the arti- ficially fed as among the breast fed infants. The influence of flies and "dirt" combined is of only slightly less importance than artificial feeding. About twice as many fly-exposed infants (2.1) in dirty homes were attacked by diarrhea as fly-protected infants in clean homes. The influence of "dirt" and artificial feeding com- bined is still greater. Three and a half times as many artificially fed infants (3.5) in dirty homes were attacked by diarrhea as breast fed infants in clean homes. Fly Factor 1.4 "Dirt" Factor 1.9 Artificial Feeding Factor 2.4 Flies and " Dirt " Factor 2.1 "Dirt" and Artificial Feeding Factor 3.5 34 TABLE I.- - NON-SELECTED NURSES. FLY PROTECTION. Total No. % Distri. Inf. Attack. by Diar. Attacks of Diar. Infants. Infants. No. % No. % Good Protection 162 37. 33 20. 43 27. Fair Protection 227 51. 25 11. 30 13. Poor Protection 52 12. 12 23. 13 25. Total 441 100. 70 16. 86 20. Fly factor 1.2 TABLE II. CONDITION OF HOME. Clean .. 126 29. 15 12. 20 16 Fair .. 121 27. 19 18. 22 18. Dirty .. 59 13. 11 19. 13 22 Unclassified .. .. 135 31. 25 19. 31 23. - - - Total.... . . 441 100. 70 16. 86 20. "Dirt" factor.. . . 1.6 TABLE III. TYPE OF FEEDING. Breast Mixed Artificial . 303 . 81 . 57 69. 18. 13. 38 14 18 13. 17. 32. 41 18 27 14. 22. 47. Total . 441 100. 70 16. 86 20. Feeding factor. . . 2.5 TABLE IV. AGE. 0-3 Months. . 128 29. 18 14. 22 17. 4-6 . 112 25. 20 18. 27 33. 7-9 . 122 28. 18 15. 21 17. 10-12 " . 79 18. 14 18. 16 20. Total . 441 100. 70 16. 86 20. Age factor.... . 1.2 TABLE V. NATIONALITY. (Grouped according to mother's.) American 112 25. 12 11. 12 11. German 18 4. 3 17. 6 33. Irish 84 19. 11 13. 14 17. Italian 101 23. 15 15. 18 18. Jew 80 18. 20 25. 25 31. 12 other Nation. 46 11. 9 20. 11 24. Total 441 100. 70 16. 86 20. 35 TABLE VI. - NON-SELECTED NURSES. FLY PROTECTION. ACCORDING TO CONDITION OF HOME. Good Protection. Total No . % Distri. Inf. Attack, by .Diar. Attacks of Diar. Infants. Infants. No. % No. % Clean . . 70 43. 9 13. 11 16. Fair . . 48 30. 10 21. 13 27. Dirty .. 11 7. 3 27. 3 27. Unclassified .. . . 33 20. 11 33. 16 49. Total.... . . 162 100. 33 20. 43 27. "Dirt" factor.. 2.1 Fair Protection. Clean . . 53 23. 5 9. 8 15. Fair .. 60 27. 7 12. 7 12. Dirty .. 25 11. 3 12. 4 16. Unclassified .. .. 89 39. 10 11. 11 12. Total.... "Dirt" factor.. . . 227 100. 25 11. . 1.3 30 13. Poor Protection. Clean 3 6. 1 33. 1 33. Fair . . 13 25. 2 15. 2 15. Dirty .. 23 44. 5 22. 6 26. Unclassified .. .. 13 25. 4 31. 4 31. Total.... "Dirt" factor.. 52 100. 12 23. . 0.7 13 25. Good, Fair and Poor Protection. Clean . . 126 29. 15 12. 20 16. Fair . . 121 27. 19 18. 22 18. Dirty .. 59 13. 11 19. 13 22. Unclassified .. .. 135 31. 25 19. 31 23. Total.. .. "Dirt" factor. .. 441 100. 70 16. .. 1.6 86 20. 36 TABLE VII. NON-SELECTED NURSES. FLY PROTECTION. ACCORDING TO TYPE OF FEEDING. Good Protection. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. No. % Attacks of Diar. No. % Breast Fed. .. ., . 102 63. 13 12. 14 14. Mixed Fed 33 20. 9 27. 13 39. Artificially Fed. 27 17. 11 41. 16 59. - - - Total 162 100. 33 20. 43 27. Feeding factor. 3.4 Fair Protection. Breast Fed 167 74. 16 10. 17 10. Mixed Fed 35 15. 2 6. 2 6. Artificially Fed. 25 11. 7 28. 11 44. Total 227 100. 25 11. 30 13. Feeding factor. . 2.8 Poor Protection. Breast Fed 34 65. 9 26, 10 29. Mixed Fed 13 25. 3 23. 3 23. Artificially Fed. 5 10. - - Total Feeding factor. 52 100. 12 23. 13 25. Good, Fair and Poor Protection. Breast Fed.. .. . 303 69. 38 13. 41 14. Mixed Fed... . 81 18. 14 17. 18 22. Artificially Fed . 57 13. 18 32. 27 47. Total . 441 100. 70 16. 86 20. Feeding factor. . 2.5 37 TABLE VIII. NON-SELECTED NURSES. FLY PROTECTION. ACCORDING TO AGE. Good Protection. Total No. % Distri. Inf. Attack. by Diar. Attacks of Diar. Infants. Infants. No. % No. % 0-3 Months. . 53 33. 12 23. 16 30. 4-6 . 38 24. 11 29. 15 40. 7-9 . 46 28. 6 13. 8 17. 10-12 . 25 15. 4 16. 4 16. Total . 162 100. 33 20. 43 27. Fair Protection. 0-3 Months. . 59 26. 4 7. 4 7. 4-6 . 67 30. 6 9. 9 13. 7-9 . 62 27. 8 13. 8 13. 10-12 . 39 17. 7 18. 9 23. Total . 227 100. 25 11. 30 13. Poor Protection. 0-3 Months. . 16 31. 2 13. 2 13. 4-6 7 13. 3 43. 3 43. 7-9 " . 14 27. 4 29. 5 36. 10-12 . 15 29. 3 20. 3 20. Total . 52 100. 12 23. 13 25. Good, Fair and Poor Protection. 0-3 Months. .. 128 29. 18 14. 22 17. 4-6 . . 112 25. 20 18. 27 24. 7-9 .. 122 28. 18 15. 21 17. 10-12 .. 79 18. 14 18. 16 20. Total . 441 100. 70 16. 86 20. 38 TABLE IX. NON-SELECTED NURSES. FLY PROTECTION. ACCORDING TO NATIONALITY. Good Protection. Total No. Infants. % Distri. Infants. Inf. Attack, bv Diar. Attacks of Diar. No. % No. % American 32 20. 5 16. 5 16. German 10 6. 1 10. 1 10. Irish 34 21. 5 15. 6 18. Italian 44 27. 8 18. 11 25. Jew 27 17. 10 37. 14 52. 12 other Nation. 15 9. 4 27. 6 40. Total 162 100. 33 20. 43 27. Fair Protection. American 72 32. 7 10. 7 10. German 7 3. 1 14. 3 43. Irish 47 21. 6 13. 8 17. Italian 43 19. 3 7. 3 7. Jew 32 14. 6 19. 7 22. 12 other Nation. 26 11. 2 8. 2 8. - - - Total 227 100. 25 11. 30 13. Poor Protection. American 8 15. - German 1 2. 1 100. 2 200. Irish 3 6. - - - - Italian 14 27. 4 29. 4 29. Jew 21 40. 4 19. 4 19. 12 other Nation. 5 10. 3 60. 3 60. Total 52 100. 12 23. 13 25. Good, Fair and Poor Protection. American 112 25. 12 11. 12 11. German 18 4. 3 17. 6 33. Irish 84 19. 11 13. 14 17. Italian 101 23. 15 15. 18 18. Jew 80 18. 20 25. 25 31. 12 other Nation. 46 11. 9 20. 11 24. - - - - Total 441 100. 70 16. 86 20. 39 June 22-27 June-July 28-4 July 5-11 July 12-18 July 19-25 July-Aug. 26-1 Aug. 2-8 Aug. 9-15 Aug. 16-22 Aug. 23-29 Aug.-Sep. 30-5 Sep. 6-12 Total Age 0-3 Months.. - 5 - 5 2 3 4 - 1 1 1 22 4-6 " - 2 3 2 4 5 4 2 3 2 - 27 7-9 1 - - 4 2 3 - 2 4 1 4 21 10-12 - - 1 1 2 2 1 5 3 1 - 16 - - - - -- - -- ■ -- - - - ■■ ■ ■ '■ Total.. .. 1 7 4 12 10 13 9 9 11 5 5 86 Total Current Attacks* of Diarrhea by Weeks and by Age Groups. Age 0-3 Months. - 5 2 6 3 5 5 3 2 2 1 34 4-6 - 2 4 3 4 5 8 9 5 3 - 43 7-9 (( 1 1 1 5 2 3 - 2 4 1 4 24 10-12 a ... - - 1 1 3 2 1 6 4 1 - 19 Total.. . 1 8 8 15 12 15 14 20 15 7 5 120 *Obtained by adding the new attacks to the attacks continued from the preceeding week. Temperatures and Relative Humidities by Weeks. Aver. Daily Mean Temp, Deg. Fahr. 73 66 69 77 73 69 73 77 77 70 75 62 Aver. Daily Mean Rei. Hum. in % 75 78 82 76 62 67 73 69 67 70 62 57 Deaths from Diarrhea of Infants under one year of age in Manhattan during summer of 1914. 17 20 24 45 51 73 77 86 98 83 77 66 TABLE X.-NON-SELECTED NURSES. Attacks of Diarrhea by Weeks and by Age Groups. 40 TABLE I. Combined Selected and Non-Selected Nurses. FLY PROTECTION. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. Attacks of Diar. No. % No. % Good Protection 416 45. 90 22. 109 26. Fair Protection 412 45. 70 17. 86 21. Poor Protection 93 10. 29 31. 35 38. Total 921 Fly factor 100. 189 21. .. 1.4 230 25. TABLE II. CONDITION OF HOME. Clean . 204 22. 29 14. 36 18. Fair . 220 24. 45 20. 56 25. Dirty . 131 14. 34 26. 40 31. Unclassified ... . 366 40. 81 22. 98 27. Total . 921 100. 189 21. 230 25. "Dirt" factor. . . 1.9 TABLE III. TYPE OF FEEDING. Breast . 636 69. 108 17. 126 20. Mixed . 182 20. 40 22. 50 27. Artificial . 103 11. 41 40. 54 52. Total Feeding factor. . 921 100. 189 21. .. 2.4 230 25. TABLE IV. AGE. 0-3 Months. . 279 30. 47 17. 56 20. 4-6 . 232 25. 50 22. 64 28. 7-9 . 240 26. 50 21. 62 26. 10-12 . 170 19. 42 25. 48 28. Total . 921 100. 189 21. 230 25. Age factor.... .. 1.5 TABLE V. NATIONALITY. American 125 13. 16 13. 16 13. Austrian 19 2. 5 26. 6 32. German 25 3. 3 12. 6 24. Irish 100 11. 14 14. 17 17. Italian 283 31. 42 15. 49 17. Jew 308 33. 95 31. 118 38. 12 other Nation. 61 7. 14 23. 18 30. Total 921 100. 189 21. 230 25. 41 TABLE VI. Combined Selected and Non-Selected Nurses. FLY PROTECTION. ACCORDING TO CONDITION OF HOME. Good Protection. Total No. Infants. % Distri. Infants. Inf. Attack. No. by Diar. % Attacks of Diar. No. % Clean . . 125 30. 19 15. 21 17. Fair . . 103 25. 23 22. 30 29. Dirty . . 39 9. 10 26. 10 26. Unclassified .. . . 149 36. 38 26. 48 32. - - - Total.... . . 416 100. 90 22. 109 26. "Dirt" factor. . 1.7 Fair Protection. Clean . . 75 18. 9 12. 14 19. Fair . . 99 24. 18 18. 21 21. Dirty . . 51 12. 11 22. 13 26. Unclassified .. .. 187 46. 32 17. 38 20. Total.... "Dirt" factor.. .. 412 100. 70 17. 1.8 86 21. Poor Protection. Clean 4 4. 1 25. 1 25. Fair .. 18 19. 4 22. 5 28. Dirty .. 41 44. 13 32. 17 41. Unclassified .. . . 30 33. 11 37. 12 40. Total.... "Dirt" factor. . .. 93 100. 29 31. 1.3 35 38. Clean Good, Fair and Poor Protection. 36 18. . . 204 22. 29 14. Fair .. 220 24. 45 20. 56 25. Dirty .. 131 14. 34 26. 40 31. Unclassified .. . . 366 40. 81 22. 98 27. Total.. .. "Dirt" factor. . . 921 100. 189 21. 1.9 230 25. 42 TABLE VII. Combined Selected and Non-Selected Nurses. FLY PROTECTION. ACCORDING TO TYPE OF FEEDING. Good Protection. Total No. Infants. % Distri. Infants. Inf. Attack, by Diar. No. % Attacks of Diar. No. % Breast Fed 282 68. 49 17. 57 20. Mixed Fed 84 20. 20 24. 24 29. Artificially Fed. 50 12. 21 42. 28 56. Total 416 100. 90 22. 109 26. Feeding factor. 2.5 Fair Protection. Breast Fed 298 72. 42 14. 48 16. Mixed Fed 69 17. 11 16. 15 22. Artificially Fed. 45 11. 17 38. 23 51. Total 412 100. 70 17. 86 21. Feeding factor. . 2.7 Poor Protection. Breast Fed 56 60. 17 30. 21 38. Mixed Fed 29 31. 9 31. 11 38. Artificially Fed. 8 9. 3 38. 3 38. Total 93 100. 29 31. 35 38. Feeding factor. . 1.3 Good, Fair and Poor Protection. Breast Fed. .. . . 636 69. 108 17. 126 20. Mixed Fed.. .. . 182 20. 40 22. 50 28. Artificially Fed . 103 11. 41 40. 54 52. Total Feeding factor . 921 100. 189 21. . . 2.4 230 25. 43 TABLE VIII. Combined Selected and Non-Selected Nurses. FLY PROTECTION. ACCORDING TO AGE. Good Protection. Total No. % Distri. Inf. Attack. by Diar. Attacks of Diar. Infants. Infants. No. % No. % 0-3 Months. . 144 35. 27 19. 33 23. 4-6 " . 98 24. 24 25. 31 32. 7-9 . 110 26. 22 20. 28 25. 10-12 . 64 15. 17 27. 17 27. Total . 416 100. 90 22. 109 26. Fair Protection. 0-3 Months. . 113 27. 16 14. 19 17. 4-6 . 116 28. 19 16. 26 22. 7-9 . 101 25. 17 17. 19 19. 10-12 . 82 20. 18 22. 22 27. Total . 412 100. 70 17. 86 21. Poor Protection. 0-3 Months. . 22 24. 4 18. 4 18. 4-6 . 18 19. 7 39. 7 39. 7-9 . 29 31. 11 38. 15 52. 10-12 . 24 26. 7 29. 9 38. - - - ■ • Total.. .. . . 93 100. 29 31. 35 38. Good, Fair and Poor Protection. 0-3 Months. .. 279 30. 47 17. 56 20. 4-6 . . 232 25. 50 22. 64 28. 7-9 . 240 26. 50 21. 62 26. 10-12 . 170 19. 42 25. 48 28. Total . 921 100. 189 21. 230 25. 44 TABLE IX. Combined Selected and Non-Selected Nurses. FLY PROTECTION. ACCORDING TO NATIONALITY. Good Protection. Total No . % Distri. Inf. Attack, by Diar. Attacks of Diar. Infants. Infants. No. % No. % American 40 10. 8 20. 8 20. Austrian 8 2. - - - ■- German 15 4. 1 7. 1 7. Irish 40 10. 8 20. 9 23. Italian 176 42. 27 15. 32 18. Jew 110 26. 37 34. 47 43. 12 other Nation. 27 6. 9 33. 12 44. Total 416 100. 90 22. 109 26. Fair Protection. American 77 19. 8 10. 8 10. Austrian 8 2. 2 25. 3 38. German 8 2. 1 13. 3 38. Irish 57 14. 6 11. 8 14. Italian 89 21. 9 10. 11 12. Jew 143 35. 41 29. 50 35. 12 other Nation. 30 7. 3 10. 3 10. Total 412 100. 70 17. 86 21. Poor Protection. American 8 9. - - - - Austrian 3 3. 3 100. 3 100. German 2 2. 1 50. 2 100. Irish 3 3. - - - - Italian 18 19. 6 33. 6 33. Jew 55 60. 17 31. 21 38. 12 other Nation, 4 4. 2 50. 3 75. Total 93 100. 29 31. 35 38. Good, Fair and Poor Protection. American .. . . ... 125 13. 16 13. 16 13. Austrian .... . . .. 19 2. 5 26. 6 32. German .... 25 3. 3 12. 6 24. Irish .... 100 11. 14 14. 17 17. Italian .... 283 31. 42 15. 49 17. Tew . ... 308 33. 95 31. 118 38. 12 other Nation. 61 7. 14 23. 18 30. Total.. . . ... 921 100. 189 21. 230 25. 45 June 22-27 June-July 28-4 July 5-11 July 12-18 July 19-25 July-Aug. 26-1 Aug. 2-8 Aug. 9-15 Aug. 16-22 Aug. 23-29 Aug.-Sep. 30-5 Sep. 6-12 Total Age 0-3 Months.. 3 7 7 7 6 6 7 5 4 3 1 56 4-6 Ci - 3 6 6 9 11 9 5 9 5 1 64 7-9 1 2 2 6 8 11 7 8 9 2 6 62 10-12 cc 2 1 3 8 6 6 4 9 6 1 2 48 Total.... 6 13 18 27 29 34 27 27 28 11 10 230 Total Current Attacks* of Diarrhea by Weeks and by Age : Groups. Age 0-3 Months. 3 9 12 14 9 11 10 13 9 6 2 98 4-6 - 3 8 8 10 16 16 14 13 6 2 96 7-9 a 1 3 4 8 9 12 10 10 12 4 7 80 10-12 a 2 3 4 8 7 7 5 11 9 2 2 60 Total.. . 6 18 28 38 35 46 41 48 43 18 13 334 *Obtained by adding the new attacks to the attacks continued from the preceeding week. Temperatures and Relative Humidities by Weeks. Aver. Daily Mean Temp, Deg. Fahr. 73 66 69 77 73 69 73 77 77 70 75 62 Aver. Daily Mean Rei. Hum. in % 75 78 82 76 62 67 73 69 67 70 62 57 Deaths from Diarrhea of Infants under one year of age in Manhattan during summer of 1914. 17 20 24 45 51 73 77 86 98 83 77 66 TABLE X.-COMBINED SELECTED AND NON-SELECTED NURSES. Attacks of Diarrhea by Weeks and by Age Groups. 46