Abstract & Commentary
Synopsis: As determined by process-of-care measurement, adverse event occurrence, and patient satisfaction, quality of care is compromised by infection control procedures.
Source: Stelfox HT, et al. JAMA. 2003;290:1899-1905.
To determine whether isolation procedures used for control of methicillin-resistant Staphylococcus aureus (MRSA) in hospitals might affect patient safety, Stelfox and associates retrospectively reviewed data from 2 large, urban teaching hospitals: Sunnybrook and Women’s College Health Sciences Centre in Toronto and Brigham and Women’s Hospital in Boston.
Stelfox et al analyzed 2 sets of patients: a consecutive series of adults admitted to the Toronto hospital for a 1-year period and a series of adult patients consecutively admitted to the Boston hospital during a 3.5-year period with a diagnosis of congestive heart failure (CHF). The latter group was studied because established standards of care relating to management of CHF facilitated an objective measurement of quality of care. In each series, patients who were managed with contact precautions, as specified by Centers for Diseases Control and Prevention guidelines, were matched to controls (2 control patients for each isolated patient) by identifying patients who occupied each isolated patient’s hospital bed immediately before and immediately after the isolated patient. Isolated patients were either colonized (in 96% of cases) or infected (in 4%) with MRSA.
Safety was assessed by 3 criteria: process of care, adverse events, and patient satisfaction. Process of care was a surrogate for thoroughness of care and included indicators such as vital sign recording, presence or absence of nurses’ and physicians’ notes, and in the CHF cohort, whether left ventricular function and ejection fraction were evaluated, whether education efforts were documented, and if follow-up appointments were scheduled. Adverse events served as a marker for outcomes of care and included injuries that lengthened hospital stay, produced disability, or resulted in abnormal laboratory test results. Patient satisfaction was assessed by identifying instances of patients’ leaving against medical advice, complaints about medical care, and altercations or suicide attempts.
The results? Isolated patients received a lower level of care, as reflected in vital sign deficiencies, absence of nurses’ and physicians’ notes, documentation of patient education and follow-up appointment scheduling, and differences in medications prescribed upon hospital discharge. In addition, isolated patients had an increased incidence of such adverse events as falls, pressure ulcers, and fluid and electrolyte disorders. Patient satisfaction was much lower in isolated patients than in controls (isolated patients were 23 times more likely to have lodged a complaint).
No differences in hospital mortality were observed.
Comment by Jerry D. Smilack, MD
This sobering, provocative report raises important questions that are infrequently asked whenever isolation procedures are instituted. In our quest to limit transmission of infections to patients and to ourselves, do we inadvertently reduce the level of care to isolated patients? What is the psychological effect of isolation on patients and their visitors?
Others1,2 have noted that health care worker contact with patients is reduced when isolation precautions are imposed. In the present study, Stelfox et al have carefully documented serious safety and medical care deficiencies associated with isolation precautions for MRSA. Since data were gathered primarily by retrospective chart review—with its attendant reliance on documentation—one might wonder what additional deficiencies would have been observed had concurrent review been in place.
Stelfox et al correctly call for further studies to determine whether certain components of isolation procedures might be more important for control of infection but less deleterious than others. They also wonder whether their findings apply to hospitals of smaller sizes or different locales.
Dr. Smilack is Infectious Disease Consultant, Mayo Clinic Scottsdale, Scottsdale, Ariz.
1. Kirkland KB, et al. Lancet. 1999;354:1177-1178.
2. Higgins L, et al. J Gen Intern Med. 2001; 16(suppl 1):200. Abstract.