From July 1, 1947 to July 1, 1949, 1251 mothers attending the obstetric clinic of the University Service of the Grace-New Haven Community Hospital were "screened" through interview to determine their preference in hospital accommodation and to obtain basic identification data. Patients were grouped according to preference for Rooming-In, preference for Nursery, and Undecided. The significance of the relationship between accommodation preference and patient's feeding preference, parity, age, education, descent, religion, race and occupation of patient's husband was then statistically determined. It was found that inclusion of the Undecided group with the Nursery group did not alter the statistical significance of any of the obtained relationships, and[SEE CHART 2 IN SOURCE PDF] consequently only the Rooming-In and Nursery groups will be contrasted in the following summary:
1. A highly significant relationship was found between accommodation preference and feeding preference. The majority of the Rooming-In group preferred breast feeding, the majority of the Nursery group preferred bottle feeding.
2. A highly significant relationship was found between accommodation preference and patient's parity. On the whole, the Rooming-In group members have lower parities than the Nursery group members.
3. No relationship was found between accommodation preference and patient's age.
4. A highly significant relationship was found between accommodation preference and occupation of patient's husband. The preponderance of patients in upper occupational classifications (student, professional, small business) occurred in the Rooming-In group. If the student classification is not included, the groups approach similarity.
5. A highly significant relationship was found between accommodation preference and patient's education. The preponderance of patients in higher educational classifications (college and above) occurred in the Rooming-In group.
6. No statistically significant relationship was found between accommodation preference and patient's descent, though a trend approaching significance was noted for the Rooming-In group to be chosen more often by patients of United States, British Isles, and Northwest European descent, and Nursery by patients of Southern European descent.
7. No relationship was found between accommodation preference and patient's religion.
8. A highly significant relationship was found between accommodation preference and patient's race. The Rooming-In group contained fewer Negro mothers in proportion to white mothers than did the Nursery group.
During the two year period, screening interviews were obtained on 280 private patients. As this group was a biased sample, the data were not analyzed statistically. Compared with clinic patients preferring rooming-in, the private patients more consistently preferred breast feeding, represented a different cultural and economic group, but were similar to the clinic group in age and parity.