Medical Care during the Depression^{: A Preliminary Report upon a Survey of Wage-Earning Families in Seven Large Cities}

The problem of giving services free has faced the doctor and the hospital during the depression to an extent unknown in any other field. A factory manager was able to economize during hard times by eliminating unproductive departments, by introducing labor-saving devices, or, as a last resort, by closing his plant until the return of prosperity. No such expedients were available to the doctor or the hospital director. Business had to continue as usual in spite of the decrease of paying patients and the tremendous increase of free care.

While the writers of this paper offer no solution for these economic problems, they do have pertinent data to present on the amount of physician's, hospital, and nursing care, both pay and free, received by a group of nearly 7,000 families in seven large cities surveyed early in 1933 by the United States Public Health Service in cooperation with the Milbank Memorial Fund. The reader is referred to previous papers3 for details, method, and scope of the survey. Briefly, it consisted of a house-to-house canvass of some 12,000 white families in the poorer districts of eight large cities, one group of coal-mining communities, and a group of cotton-mill villages. The records obtained by the canvasses included (a) the economic history of the family in sufficient detail for computing family income for each year from 1929 through 1932, and (b) a record of all illness during the three months immediately preceding the date of the enumerator's visit, in the spring of 1933, with the extent of disability and of medical care for each case.

The sample population discussed in the present paper comprised 28,959 individuals in 6,686 families for which the data were sufficiently complete for computing the actual income for each of the four years from 1929 to 1932. The population was largely of the wage-earning class, a considerable proportion of which had experienced loss of income due to unemployment and wage reductions. In 1929, 10 per cent of the persons surveyed were in families with an annual per capita income of $149 or less; by 1932, 43 per cent were in this class. On the other side of the picture, 42 per cent of the persons were in families with an annual per capita income of $425 or more in 1929, but by 1932 this figure had decreased to 14 per cent.

Medical Care in a Surveyed Group

Tables 1 and 2 summarize the data for the entire group. We see that 52.4 per cent of all cases of illness received attendance of some kind during the three-month survey period; 67.7 per cent of disabling4 illnesses and 30.0 per cent of non-disabling illnesses were attended. Attendance by a physician5 accounted for the greater part of the care received—51.7 per cent of all illnesses were attended by a physician and 40.8 per cent had no other service except that of a physician. Expressed differently, 99 per cent of the illnesses that received care of any sort had a doctor and in 78 per cent of the cases a doctor was the only attendant, the other 21 per cent having hospital or nursing care in addition to a physician. Considering disabling illness, 67 per cent received the care of a physician and in 49 per cent the doctor was the only attendant.

TABLE 1.

Extent of Medical Care—Per Cent of Total Illnesses and of Disabling and Non-Disabling Illnesses Receiving Medical, Hospital, and Nursing Services during a Three-Month Period in 1933¹

	Per Cent of Illnesses Receiving Specified Services
Service	All Illness	Disabling Illness	Non-Disabling Illness
Any service	52.4	67.7	30.0
Physician	51.7	66.9	29.4
Physician only	40.8	49.1	28.9
Physician and hospital	8.4	14.2	—
Physician and visiting nurse	2.2	3.3	0.5
Physician and bedside nurse	0.1	0.2	—
Hospital	8.4	14.2	—
Excl. of cases hospitalized 90 days	7.4	12.4	—
Visiting nurse	3.8	5.8	1.1
Visiting nurse only	0.7	0.9	0.5
	Illness Rate per 1,000 Persons
Illness rates	237	141	96

¹Based on 28,959 individuals in 6,686 wage-earning families surveyed in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse.

TABLE 2.

Kinds of Medical Care Received—Distribution of Illnesses Which Received Medical Care according to Kind of Services Received during a Three-Month Period in 1933¹

	Per Cent Receiving Specified Service
Service	Total	Disabling Illness	Non-Disabling Illness
Any service	100.0	100.0	100.0
Physician	98.6	98.7	98.1
Physician only	77.8	72.4	96.3
Physician and hospital	16.1	20.9	—
Physician and visiting nurse	4.2	4.9	1.8
Physician and bedside nurse	0.2	0.3	—
Hospital	16.1	20.9	—
Excl. of cases hospitalized 90 days	14.0	18.3	—
Visiting nurse	7.4	8.5	3.6
Visiting nurse only	1.4	1.3	1.6

¹See footnote to Table 1.

Of all illnesses, 8.4 per cent had hospital care within the three-month survey period and of all disabling illnesses, 14.2 per cent had such care. Excluding cases in hospitals during the entire ninety days of the survey period, principally patients in public mental and tuberculosis sanitariums, 7.4 per cent of all illnesses and 12.4 per cent of disabling illnesses received hospitalization. Attendance by a visiting nurse was received by 3.8 per cent of the illnesses; 2.2 per cent had both visiting nurse and physician within the three-month survey period. The group received a negligible amount of care by a bedside nurse and hence this service is not considered in the tables that follow.

Comparison with Results of Other Surveys

Comparison with the data of the Committee on the Costs of Medical Care6 would indicate that the group of wage-earning families here considered received less total care than the lower income groups of that study, $3,000 and under, which correspond most nearly to the survey group herein discussed. In the Committee's group, 66.5 to 80.4 per cent of illnesses during a period of one year received service of some kind (the larger part of this being services of a physician) as compared with our figure of 52.4 per cent. Hospital care, however, is about the same in both surveyed groups—6.6 to 7.4 per cent for the Committee's survey and 7.4 per cent for the present group when cases with ninety days in the hospital (the whole survey period) are excluded.7

A survey of the Metropolitan Life Insurance Company8 in 1915–1917, recording the illnesses among some 600,000 persons on the day of the canvass, indicated that 9.9 per cent of the persons sick and unable to work were in the hospital. This figure varied from 3.0 per cent in North Carolina to 19.3 per cent in Boston; the combined data for the cities of Boston, Kansas City, New York, Pittsburgh, and Trenton give a figure of 13.1 per cent. The proportion of disabling illnesses hospitalized, 14.2 per cent, in the present survey is not far different from these figures of the Metropolitan Life Insurance Company survey.9

Sydenstricker,10 in a study of the incidence of illness in Hagerstown from December 1, 1921 to March 31, 1924, found 1.3 per cent of the cases hospitalized and 46 per cent attended by a physician. These figures on the extent of hospitalization are much lower than for surveys in larger cities; the attendance by a physician is not far from the figure obtained in the present survey (52 per cent) but lower than the figure of the Committee on the Costs of Medical Care (67 to 80 per cent).

To summarize, comparison of the results of the present study with those of other surveys indicates that the canvassed group received as much hospitalization as is customary for people in these economic classes but probably less care by a doctor. Internal comparisons in the group, as will be shown later, point to the same conclusion.

Illness and 1932 Income

Before discussing economic status and the care received for illness, the incidence of illness in the different economic groups will be considered briefly. Family income per capita has been used as a measure of the well-being of the family. For convenience in discussion, the groups are designated as follows: “poor”—under $150 per capita per year; “moderate”—$150–$424 per capita per year; “comfortable”—$425 and over per capita per year. Figure 1 shows illness rates for the canvassed population classified in the foregoing three income groups.

Figure 1.

Disabling and Non-Disabling Illness during a Three-Month Period in the Early Spring of 1933 in Wage-Earning Families Classified according to per Capita Income in 1932 in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse

Considering disabling illnesses, onset within and prior to the survey period, the “poor” group shows an illness rate 22 per cent higher than the “comfortable” group—152 as against 125 cases per 1,000 persons. Non-disabling illness rates show no apparent association with income. The differences in illness rates are largely due to differences in illnesses having their onset within the survey period; the cases with prior onset (principally chronic) show little change with economic status. Hence percentage difference is greatest when disabling illnesses, onset within the study, are considered. The “poor” group show a rate for these acute disabling illnesses of 108 cases per 1,000 persons which is 35 per cent higher than the rate of the “comfortable” group, 80 cases per 1,000 persons.

Units of Measurement and Basic Results

Tables 3 and 4 give in some detail the attendance for illness by physician, hospital, and visiting nurse in three groups of the surveyed population classified by per capita income in 1932. The per cent of all illnesses receiving the specified service is shown in Table 3, and the volume of service, that is, physician's or nurse's calls or days of hospital care, is shown in Table 4 in two ways, (a) per 1,000 persons under observation, and (b) per 1,000 cases of all illness. The illness figures used as the base for these rates are all illnesses, whether or not care was received, disabling or non disabling, with onset prior to or within the study.

TABLE 3.

Income and Medical Care—Per Cent of Total Illnesses Receiving Medical, Hospital, and Nursing Services Related to 1932 Family Income per Capita, in Canvassed White Families in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse¹

	Per Cent of Illnesses Receiving Specified Service
Service in Specified per Capita Income Groups2	Total Care	Pay Care	Free Care
Physician
Poor	50.2	18.8	31.4
Moderate	51.4	34.3	17.1
Comfortable	58.0	45.9	12.1
Hospital, all cases
Poor	9.5	1.3	8.2
Moderate	7.8	2.8	5.0
Comfortable	6.9	3.4	3.5
Hospital, excl. of cases hospitalized 90 days
Poor	8.4	1.2	7.2
Moderate	6.8	2.8	4.0
Comfortable	6.0	3.3	2.7
Visiting nurse
Poor	5.6	0.1	5.5
Moderate	2.8	0.1	2.7
Comfortable	1.2	0.2	1.0

¹The illness and population figures on which Tables 3 and 4 are based are as follows:

²Poor—under $150 per capita per year.

Moderate—$150–$424 per capita per year.

Comfortable—$425 and over per capita per year.

TABLE 4.

Income and Volume of Medical Service—Physician's or Nursing Calls or Days of Hospital Care per 1,000 Persons in the Canvassed Population and per 1,000 Illnesses (Disabling and Non-Disabling) Related to 1932 Family Income per Capita in Canvassed White Families in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse¹

	Volume of Service, Calls, or Days
	Per 1,000 Persons			Per 1,000 Illnesses
Service in Specified per Capita Income Groups2	Total Care	Pay Care	Free Care	Total Care	Pay Care	Free Care
Physician
Poor	558	219	339	2,219	869	1,350
Moderate	677	456	221	2,963	1,998	965
Comfortable	817	630	187	3,699	2,852	847
Hospital, all cases
Poor	575	72	503	2,293	287	2,006
Moderate	447	85	362	1,963	375	1,588
Comfortable	371	115	256	1,681	524	1,157
Hospital, excl. of cases hospitalized 90 days
Poor	323	43	280	1,304	173	1,131
Moderate	233	79	154	1,031	348	683
Comfortable	187	93	94	855	423	432
Visiting nurse
Poor	79	3	76	319	14	305
Moderate	48	2	46	207	7	200
Comfortable	14	5	9	62	23	39

¹See footnote to Table 3.

²See footnote to Table 3.

Table 5 gives similar data for the sample population grouped according to change in economic status from 1929 to 1932. No attempt will be made here to discuss the data in detail. Two graphs, Figures 2 and 3, show that the relations among the data are similar whatever base is employed. The units used in the graphs are the volume of care, expressed as calls or days, received per 1,000 cases of all illnesses whether attended or unattended. The volume of service per 1,000 cases of illness is used rather than the volume per 1,000 persons under observation, as it eliminates the effect of varying illness rates in the different groups.

TABLE 5.

Income Change, 1929–1932, and Physician's and Hospital Service in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse

			Volume of Service, Calls, or Days
Economic Status1 in		Per Cent of Illnesses Receiving Service			Per 1,000 Persons			Per 1,000 Illnesses
1929	1932	Total Care	Pay Care	Free Care	Total Care	Pay Care	Free Care	Total Care	Pay Care	Free Care	Number of Cases2	Population Observed
I. Physician's Care
Poor	Poor	48.7	13.5	35.2	545	133	412	2,283	555	1,728	688	2,884
Moderate	Poor	52.2	20.1	32.1	529	223	306	2,241	947	1,294	1,675	7,109
Moderate	Moderate	50.0	31.2	18.8	655	424	231	2,998	1,938	1,060	1,342	6,139
Comfortable	Poor	47.1	20.7	26.4	651	302	349	2,113	980	1,133	774	2,513
Comfortable	Moderate	52.4	36.7	15.7	697	490	207	2,946	2,069	877	1,445	6,101
Comfortable	Comfortable	58.6	46.4	12.2	834	641	193	3,745	2,879	866	817	3,672
II. Hospital Care3
Poor	Poor	10.6	0.9	9.7	358	29	329	1,531	126	1,405	676	2,884
Moderate	Poor	8.1	1.4	6.7	290	45	245	1,243	195	1,048	1,658	7,109
Moderate	Moderate	5.8	2.4	3.8	180	59	121	833	274	559	1,323	6,139
Comfortable	Poor	6.7	1.0	5.7	377	51	326	1,234	168	1,066	767	2,513
Comfortable	Moderate	7.4	3.1	4.3	263	99	164	1,118	419	699	1,435	6,101
Comfortable	Comfortable	5.9	3.1	2.8	196	95	101	886	430	456	811	3,672
III. Visiting Nurse's Care
Poor	Poor	7.0	0.5	6.5	92	7	85	387	28	359	688	2,884
Moderate	Poor	5.0	0.0	5.0	78	0	78	332	0	332	1,675	7,109
Moderate	Moderate	2.8	0.0	2.8	33	0	33	152	0	152	1,342	6,139
Comfortable	Poor	5.8	0.1	5.7	71	9	62	231	31	200	774	2,513
Comfortable	Moderate	2.5	0.1	2.4	46	3	43	196	13	183	1,445	6,101
Comfortable	Comfortable	1.2	0.2	1.0	13	5	8	62	24	38	817	3,672

¹Poor—under $150 per capita per year; moderate—$150–424 per capita per year; comfortable—$425 and over per capita per year.

²Number of cases of all illness whether or not attended. Case rates per 1,000 persons for the three-month period are as follows; poor-poor, 239; moderate-poor, 236; moderate-moderate, 219; comfortable-poor, 308; comfortable-moderate, 237; comfortable-comfortable, 222.

³Excludes cases hospitalized 90 days (the entire survey period) which were largely patients in public mental and tuberculosis sanitariums.

Figure 2.

Service per Case, Total Disabling and Non-Disabling, by Physician, Hospital, and Visiting Nurse during a Three-Month Period in the Early Spring of 1933 in Wage-Earning Families Classified according to per Capita Income in 1932 in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse

Figure 3.

Service per Case, Total Disabling and Non-Disabling, by Physician, Hospital, and Visiting Nurse during a Three-Month Period in the Early Spring of 1933 in Wage-Earning Families Classified according to Change in per Capita Income, 1929–1932, in Baltimore, Birmingham, Cleveland, Detroit, New York, Pittsburgh, and Syracuse

Medical Care and 1932 Income

Figure 2 shows two different sequences with increasing economic well-being, (1) the total service by physicians increases, and (2) the service by hospitals and visiting nurses decreases. The “poor” evidently get more hospital care and more calls by visiting nurses than the “moderate” and “comfortable” but fewer physicians’ calls.

Considering first the care by physician, the “poor” received 2,219 calls per 1,000 total illnesses attended or unattended, and the “comfortable” 3,699 calls, or 67 per cent more care. This difference was entirely due to the greater amount of care paid for by the “comfortable” class; this was nearly three and one-half times that of the “poor” group—2,852 calls per 1,000 illnesses as compared with 869. The “poor” group received more than one and one-half times the free care that the “comfortable” group received. In percentages of the total, 61 per cent of all physicians’ calls to the “poor” were free, as compared with 33 per cent for the “moderate” and 23 per cent for the “comfortable.”

Considering hospital service exclusive of cases hospitalized the whole ninety days, the “poor” group received 1,304 days care per 1,000 illnesses (attended or unattended); the “moderate” 1,031; and the “comfortable” 855. Thus the “poor” received a 53 per cent greater volume of care than the “comfortable.” The “comfortable,” however, paid for more hospital care than the “poor;”“comfortable,” 423 days, “poor” 173 days per 1,000 total illnesses—or about two and one-half times as much.

Calls by a visiting nurse were practically all free; the “poor” group received 319 calls per 1,000 cases of illness (attended or unattended) as compared with 62 for the “comfortable” group.

Medical Care and Change in Income

The years between 1929 and 1932 witnessed tremendous changes in family income, largely in a downward direction. Not all of the families that were poverty-stricken in 1932 were accustomed to this misfortune. Considering the 12,500 individuals in families classified as “poor” (under $150 per capita income) in 1932, 23 per cent were poor in 1929, 57 per cent were in moderate circumstances ($150–424 per capita income) in the earlier year, and 20 per cent were classified as comfortable ($425 and over per capita income). It is of interest to examine the medical and hospital care received by groups of individuals classified according to economic status in 1929 and in 1932. This has been done in Figure 3. Here, for example, the “poor” group in 1932 is now divided into the “chronic poor” who were poor in 1929 and 1932, and two groups of the “depression poor”—those who were “comfortable” in 1929 but “poor” in 1932 and those who were in the “moderate” class in 1929 but were “poor” in 1932. For the whole group which was “poor” in 1932 (Figure 2) there were received 2,219 physicians’ calls per 1,000 cases of illness, of which 39 per cent was pay and 61 per cent free care. From Figure 3, we see that the total amount of physicians’ care was about the same for the “chronic poor” as for the newly “poor.” However, the “poor” who had been “poor” even in 1929 paid for only 24 per cent of the physicians’ calls received while the “poor” who had been in the “moderate” class in 1929 paid for 42 per cent and the “poor” who had been in the “comfortable” class paid for 46 per cent of the total calls received.

Considering hospital care, the whole group which was “poor” in 1932 received 1,304 days per 1,000 cases of illness (hospitalized or nonhospitalized). In Figure 3, it is seen that the “chronic poor” received more hospital service than the “poor” who had been in better circumstances in 1929—1,531 days per 1,000 cases as compared with about 1,250 days per 1,000 cases for both the “comfortable-poor” and “moderate-poor” groups. Of the hospital service received by the “chronic poor,” 92 per cent was free as compared with 85 per cent in both classes of the “depression poor.” Apparently the “new poor” had not made as good a connection with sources of free care as those who had been in straitened circumstances for a longer period of time.

Care by a visiting nurse showed much the same picture as the hospital care (Fig. 3).

For further comparisons, we may assume that families which showed little change in economic status between 1929 and 1932 obtained in 1932 about the customary11 amount of medical care for individuals of their income level and social status. With this idea in mind, we may compare, for example, the group “comfortable” in 1929 and 1932 with the less fortunate group which was in similar circumstances in 1929 but was reduced to poverty by 1932—the “comfortable 1929–poor 1932” group.

It is seen from Figure 3 and Table 5 that the “comfortable-comfortable” received a total of 3,745 physicians’ calls per 1,000 cases of illness, a volume of service almost twice that of the “comfortable-poor” group which received 2,113 calls per 1,000 illnesses. A similar comparison shows that the “moderate-moderate” group received more calls per 1,000 illnesses (2,998) than the “moderate-poor” (2,241).

Free care constituted 23 per cent of physicians’ calls to the “comfortable-comfortable” and 54 per cent of the calls to the “comfortable-poor”; free care was 35 per cent of physicians’ calls to the “moderate-moderate” as compared with 58 per cent of the calls to the “moderate-poor.”

The groups which suffered no income reduction during the depression obtained twice to three times the volume of paid physicians’ calls and about 20 per cent less free calls than their less fortunate neighbors who suffered heavy reverses. The net result was the receipt of considerably more doctor's care by families with unchanged income.

For hospital care, a different relation is evident. The groups which had remained in the “comfortable” or “moderate” class throughout the four years received less total hospital care than the groups of the “depression poor.” For example, the “comfortable-comfortable” show 886 days hospital care per 1,000 cases of illness which is about one-third less volume of care than received by the “comfortable-poor” (1,234 days per 1,000 illnesses). The “moderate-moderate” also received about one-third less volume of hospital service than the “moderate-poor”—833 days as compared with 1,243 days per 1,000 illnesses.

Free care was 51 per cent of total hospital days for the “comfortable-comfortable”; 86 per cent for the “comfortable-poor”; 67 per cent for the “moderate-moderate”; and 84 per cent of the “moderate-poor.”

The smaller amount of hospital care received by the classes that suffered little change in economic status is entirely due to the small amount of free care received by these groups as compared with the groups reduced to the poverty level during the depression.

Thus, internal comparisons among various groups of the surveyed population indicate that the “depression poor” obtained more free care of all kinds, less total physicians’ care, and more total hospital care and care by a visiting nurse than was received by their neighbors who were in similar economic circumstances in 1929 but did not suffer material loss of income during the depression.

Summary

This paper presents a preliminary analysis of hospital, nursing, and physician's care received by wage-earning families severely affected by the depression. The data were obtained as part of a house-to-house sickness survey in seven large cities. The results indicate that a very large proportion of the total service received by the group was free. The volume of this free care in various groups of the population classified by income was from 25 to 75 per cent of the physician's calls, 50 to 90 per cent of the hospital days, and 60 to 100 per cent of the calls by a visiting nurse. The “chronic poor,” a group which were poverty stricken even in 1929, show the largest percentage of free care and the largest total volume of hospital and visiting nurses’ service. Families that had suffered loss of income during the depression (the depression poor) received more hospital care, largely free, than families of similar economic status in 1929 that had not lost income. Total care by a physician was less among the poor than among the moderate and comfortable, but here again the poor received more free care. Internal comparisons among the different economic groups indicate that families reduced to poverty between 1929 and 1932 received more free care of all kinds, more total service by hospitals and visiting nurses, and less physician's care than was received by families which remained in moderate or comfortable circumstances throughout the economic depression.