Page:LewisMeriam-TheProblemOfIndianAdministration.djvu/295

in the Navajo country. In the preparation and analysis of the statistics the Indian Office must recognize these facts and separate its accurate reliable figures from the less accurate. To combine accurate and inaccurate figures in inseparable totals may lead to erroneous conclusions. Accurate facts are diluted with inaccurate. Separate figures should be given even if this course necessitates dividing jurisdictions into districts according to the reliability of the data for them. The Office is now beginning to recognize this principle and is asking that the distinction be made.

Factors Affecting Health Statistics. Other factors that affect the accuracy and completeness of health statistics which must be considered are:

1. Returns made by Field Personnel other than the Physician: The various field workers on the reservations have been instructed to report all births and deaths coming to their attention. Frequently such reports are made to the agency office and not to the physician. Such records relating to deaths are almost certain to be incomplete and inaccurate because at best the diagnosis as to cause of death is a guess. Often no report at all as to cause is given. Several such instances were found. At Fort Peck in the past five years, 53.2 per cent of the deaths recorded in the office were reported without a statement of the cause. At Pine Ridge in the past eighteen months 31.2 per cent had no statement of cause of death. At Cheyenne River in a period of four years and three months 37 per cent were reported without cause of death. On the other hand at Crow Agency where three physicians and three field matrons serve an Indian population of approximately 1800, during a period of three years a definite cause of death was reported for all but 2.8 per cent of the deaths. The Consolidated Chippewa Agency in Minnesota had a record of the cause of death in all but 10.8 per cent in a four-year period. The Rosebud Reservation, comprising approximately 5700 Indians scattered over an area of forty by sixty miles, with two full time and two contract physicians and a public health nurse, reported a definite cause of death in all but 13.5 per cent of the cases.
2. Lack of Accurate Case Records: Indian Service physicians have never kept accurate case records of disease and illness in Indian families. Thus the transferred physician must learn his new field from contact with cases, a thing which would require an enor-