Homesickness and Children's Adjustment to Hospitalization: Toward a Preliminary Model

To better understand the role of homesickness in children's adjustment to hospitalization, we culled self-report and observational data from hospital staff, primary caregivers, and children (n = 50) in a variety of inpatient units (psychiatric, burn, rehabilitation). Homesickness was more severe and less predictable than in comparable samples of children in less stressful environments (e.g., summer camp). The best predictors were negative hospitalization attitudes and previous separations from home, such as foster placements. Contrary to our hypotheses, insecure attachment and low perceived control did not predict homesickness. In the context of a “homesick disposition” model, prevention strategies should emphasize information provision, positive modeling, and coping skills training.

ACKNOWLEDGMENTS

This research was supported by a National Institutes of Health postdoctoral training grant in rehabilitation medicine awarded to Marjorie Anderson at the University of Washington School of Medicine. Manuscript preparation was completed while Christopher A. Thurber was supported by the Colin Brown Memorial Fund. Family and staff incentives were supported by generous donations from the McDonald's Corporation and Tully's Coffee.

We wish to thank Drs. Marjorie Anderson, Kenneth Jaffee, Rose Calderon, Elizabeth McCauley, David Heimbach, Loren Engrav, and Art Lewy for their cooperation and creative input. We also extend our gratitude to the participating children and their families who gave unselfishly of their precious time.

Notes

∗Note.    Superscript numbers refer to the following key references and review articles: 1. Prugh, Staub, Sands, Kirschbaum, & Lenihan (1953); 2. Langford (1961); 3. Vernon, Foley, Sipowicz, & Schulman (1965); 4. Tiedeman & Clatworthy (1990); 5. Lavigne & Burns (1981); 6. Peterson, Mori, & Carter (1985); 7. Rutter (1972, 1979); 8. Carson, Council, & Gravley (1991); 9. McClowry & McLeod (1990); 10. Fritsch, Heinssen, Delga, Goodrich, & Yates (1992); 11.Robinson, Rankin, & Drotar (1996); 12. Ziegler & Prior (1994); 13. Chapman, Loeb, & Gibbons (1956); 14. Saylor et al. (1987); 15. Fosson, Martin, & Haley (1990); 16. Mabe, Treiber, & Riley (1991); 17. Dalton, Daruna, & Bolding (1991); 18. Kim, Hahn, Kish, Rosenberg, & Harris (1991); 19. Belson (1993); 20. Turner, Tomlinson, & Harbaugh (1990); 21. Thompson (1985); 22. Douglas (1975); 23. Brian & Maclay (1968); 24. Sides (1977); 25. Skipper & Leonard (1968); 26. Blader, Abikoff, Foley, & Koplewicz (1994); 27. Whelan & Kirkby (2000); 28. Whelan & Kirkby (1998); 29. Hazzard, Celano, Collins, & Markov (2002).

¹The use of the CBCL to assess children with chronic illnesses has been discouraged by some authors (e.g., LaGreca, 1994; Perrin, Stein, & Drotar, 1991). In this case, it was the best parent-report measure of prehospitalization functioning because of its validity for other types of hospitalized children and because it permitted direct comparison with prior studies of homesick children, most of which had used the CBCL.

²Other staff-report measures commonly used to measure adjustment in hospital settings include the Behavioral Upset in Medical Patients—Revised (BUMP–R; Rodriguez & Boggs, 1994) and the Pediatric Inpatient Behavior Scale (PIBS; Kronenberger, Carter, & Thomas, 1997). However, neither was ideally suited for the daily staff reporting demands of this study because both are lengthy—the BUMP–R is 56 items and the PIBS is 47 items—and neither assesses homesickness or symptoms of separation anxiety.

³The most conservative cutoff score of 19 for the CDI was maintained, despite the fact that one item—that on suicidality—was not administered, thus lowering the possible maximum score to 52 from 54. The cutoff of 19 was chosen for consistency with previous research. Had the cutoff been lowered to 17, the CDI would have categorized 12 children as severely depressed, but would not have changed the overall homesickness results significantly.

∗p < .05, two-tailed.

∗∗p < .01, two-tailed.

∗p < .05, two-tailed.

∗∗p < .01, two-tailed.

∗p < .05, two-tailed.

∗∗p < .01, two-tailed.