Divergent fates of the medical humanities in psychiatry and internal medicine: should psychiatry be rehumanized?

Abstract

Objective: To determine the degree to which the medical humanities have been integrated into the fields of internal medicine and psychiatry, the authors assessed the presence of medical humanities articles in selected psychiatry and internal medicine journals from 1950 to 2000.

Methods: The journals searched were the three highest-ranking psychiatry and internal medicine journals on the Institute for Scientific Information's Impact Factor rankings that were published in English and aimed at a clinical audience. Operationalized criteria defining the medical humanities allowed the percentage of text in the selected journals constituting medical humanities to be quantified. Journals were hand searched at 10-year intervals from 1950 to 2000. Mixed effects models were used to describe the change in medical humanities over time.

Results: The percentage of text within psychiatry journals meeting the criteria for medical humanities declined from a peak of 17% in 1970 to a low of 2% in 2000, while the percentage of humanities articles in internal medicine journals roughly doubled from 5% to 11% over the same time period. A linear model increasing over time best fit the medical humanities in the internal medicine journals, while a cubic model decreasing over time best fit the psychiatry humanities data. Humanities articles in medical journals had a greater breadth and diversity than those in psychiatry journals.

Conclusion: Medical humanities publications dramatically decreased over time in psychiatry journals while they more than doubled in internal medicine journals. These data suggest the need for further empirical research and discussion of the potential roles of the humanities in psychiatry.

FIGURE 1.

Percent of Publication Pages Devoted to Medical Humanities in Selected Psychiatry and Internal Medicine Journals from 1950 to 2000

FIGURE 2.. Comparison of Average Percent of Publication Pages Devoted To “Medical Humanities and “Arts” in Selected Psychiatry and Internal Medicine Journals from 1950 to 2000

^aData for internal medicine were best fit using a linear regression model, (R²=0.697, p=0.039), with an increasing trend over time. Data for psychiatry were best fit using a cubic model (R²=0.99, p=0.001), with a decreasing trend over time. ^bData for internal medicine fit linear (R²=0.664, p=0.048, increasing over time) and exponential (R²=0.729, p=0.03, increasing over time) models. Data for psychiatry fit a liner regression model with negative slope only at the trend level (R²=0.58, p=0.078).