Current issues of ACP Journal Club are published in Annals of Internal Medicine


Physical therapy led to modest improvements in mobility in very frail nursing home residents

ACP J Club. 1994 July-Aug;121:15. doi:10.7326/ACPJC-1994-121-1-015

Related Content in the Archives
Review: Multiple risk factor modification reduces falls in elderly persons

Source Citation

Mulrow CD, Gerety MB, Kanten D, et al. A randomized trial of physical rehabilitation for very frail nursing home residents. JAMA. 1994 Feb 16;271:519-24.



To assess whether physical therapy (PT) for very frail long-stay nursing home residents improved physical function and self-perceived health.


4-month randomized controlled trial.


9 nursing homes in Texas.


194 nursing home residents (mean age81 y, 70% women) who were > 60 years old, resided in the nursing home for ≥ 3 months, and were dependent in ≥ 2 activities of daily living (ADL). Exclusion criteria were terminal illness or medical condition precluding participation in PT, severe dementia (a prorated score of < 50% on the Folstein Mini-Mental State Examination [MMSE]), inability to follow a 2-step command, assaultive behavior, or recent or ongoing PT. 180 participants (93%) completed the study.


Participants were stratified by nursing home and were randomly assigned to receive PT (n = 97) or friendly visits (n = 97). Participants receiving PT had 1-on-1 sessions, 3 times/wk, with 1 of 6 physical therapists who provided a standardized assessment and individualized therapy selected using a prioritized algorithm. Friendly visits occurred 3 times/wk and usually entailed reading to participants. Compliance was defined as the proportion of scheduled sessions completed.

Main outcome measures

Physical status assessed by the Physical Disability Index (PDI), general function assessed by the Katz ADL Scale, self-perceived health assessed by the Sickness Impact Profile, cognition and depression levels assessed by the Geriatric Depression Scale (GDS) and the Folstein MMSE, and the number of falls.

Main results

No improvement in any outcome measure was noted except for the mobility subscale of the PDI, which showed an improvement of 15.5% (95% CI 6.4% to 24.7%, P = 0.01) in the group receiving PT compared with the group receiving friendly visits. The PT group was less likely to use assistive devices and wheelchairs (P < 0.005) but tended to fall more often (P = 0.11). Average charge per participant for PT was $1220 (CI $412 to $1832) with no difference in other health care charges for the 2 groups. Posterior power calculations showed a power of 80% to detect changes of ≤ 20% between groups for all outcome measures except the GDS, which showed a power of 80% to detect changes of 24%.


Modest improvements in mobility were found among very frail long-stay nursing home residents after a 4-month physical therapy program. Program costs were substantial.

Sources of funding: National Institute on Aging and Veterans Affairs Health Services Research and Development.

For article reprint: Dr. C.D. Mulrow, Audie L. Murphy Memorial Veterans Hospital, Ambulatory Care (11C), 7400 Merton Minter Boulevard, San Antonio, TX 78284, USA. FAX 210-567-4685 or 210-567-4423.


The Omnibus Budget Reconciliation Act of 1987 mandated increased attention to the quality of life in nursing homes and required that all residents receive services designed to maximize their function. Although long-stay residents account for most of the costs and have the greatest needs, few receive formal rehabilitation. The largest barriers have been a lack of funding (Medicare is limited to short-term rehabilitation) and a lack of proven efficacy.

In a well-designed study with sufficient power to detect meaningful improvements, Mulrow and colleagues provide data that show that PT can be beneficial, but their results are unlikely to convince policy-makers to expand funding. The mobility gains were modest and may not have been welcomed by those patients who had more musculoskeletal pain and falls or by staff who had to increase their supervision of a more mobile but still highly impaired group. Therapeutic gains are potentially short-lived. My own experience suggests that functional improvement in this population often reverses after therapy ceases, particularly if acute illness occurs or other aspects of interdisciplinary care are unchanged. Finally, the costs of PT, which represented 4% of the yearly total cost of institutional care, would be substantial if provided universally.

Nursing homes are "total institutions," whose very environment encourages physical dependency by valuing bodily safety and regimentation more than personal choice (1). Patients often lack the motivation, self-efficacy, or attention span to participate in focused therapy, and their physical needs do not end when a formal program ceases. Cost-effective restorative care may require re-inventing the nursing home, moving away from a medical model toward providing activities that patients actively choose and enjoy.

Joseph Francis, MD
University of TennesseeMemphis, Tennessee, USA


1. Lidz C, Fischer L, Arnold RM. The erosion of autonomy in long-term care. New York: Oxford University Press; 1992.