Poor nutrition and dehydration are common in nursing home residents and are associated with many adverse clinical outcomes. OBRA (Omnibus Budget Reconciliation Act) guidelines require assessment if the resident is documented to be eating 75 percent or less of most meals. The nursing home staff typically makes this determination. However, studies reveal that discrepancies exist between the estimates and the actual nutritional intake. Simmons and Reuben conducted a validation study to compare three methods of assessing food and fluid intake by nursing home residents.

Nursing home residents who were more than 65 years of age, who understood English and were incontinent but free of a catheter were included in the study. A Folstein Mini-Mental State Examination (MMSE) was conducted to assess the cognitive function of each participant. Total percentage of food and fluid intake of each participant for each of nine meals (three meals per day over three consecutive days) was assessed by three methods: (1) nursing home staff chart documentation; (2) research staff documentation, based on direct observation; and (3) research staff documentation based on pre- and post-meal photographs. A researcher with a master's degree in nutrition then determined, based on the pre- and post-meal photographs, the percentage of food and fluid consumed by each participant.

The average age of the 56 participants included in the study was 86.2 years, and a broad range of cognitive status was represented within the sample, as evidenced by MMSE scores ranging from zero to 26, with an average of 11.7. Nursing home staff estimated a nutritional intake of 79.3 percent, direct observation by the trained researchers estimated a nutritional intake of 58.9 and estimation by the researcher assessing pre- and post-meal photographs was 58.2 percent. The differences between the nursing home staff estimates and estimates by the two research staff methods were statistically significant. Nursing home staff documentation revealed a significant intake overestimation of 22 percent compared with the two research staff methods. The lower the percentage of intake for a particular resident, the higher the difference between nursing home staff documentation and the photography method. Research staff direct observation and research staff photography results were similar and, in addition to the total percent intake for each meal, the two research methods yielded the percent intake for individual food and fluid items. Significant agreement between the research staff methods for all types of food and fluid items across all three meals was evident. The three methods were compared in identifying patients whose total nutritional intake met the OBRA guidelines. This comparison revealed that nursing home staff failed to identify 19 of 36 patients (53 percent) who were identified by the two research methods as meeting OBRA guidelines for further assessment.

The authors conclude that nursing home staff overestimate the nutritional intake of residents by as much as 22 percent—a photographic record is a possible solution to ensure accurate assessment. Disposable cameras were used in this study at a cost of approximately $630 to monitor nine meals for 56 nursing home residents. However, the study did not determine research staff time required to implement each of the three methods, which is also an important cost factor. The photographic method could be used for selected residents (e.g., those at risk for undernutrition or dehydration), as a random sampling technique to help the nursing home staff better document nutritional intake levels and as a training tool. The photographic method is a reliable and time-efficient method of assessing nursing home residents' nutritional intake. Further studies of its cost-effectiveness may be needed.