Design
This study retrospectively compared pre- and post-intervention incidences of pneumonia.
Setting and subjects
Almost all patients with behavioral and psychological symptoms of dementia (BPSD) are treated in psychiatric hospitals in Japan, and quite a few patients with dementia remain there long-term, even after BPSD are well controlled. Therefore, many dementia patients die in psychiatric hospitals in Japan.
We enlisted the member hospitals of the Association of Okayama Psychiatry Hospitals in a survey of inpatients undergoing artificial feeding. Of 20 psychiatric hospitals in Okayama Prefecture, three do not care for patients with dementia or psychiatric diseases in their chronic or terminal state. Nine of the remaining 17 agreed to participate in this survey.
All patients fulfilled following criteria. (i) They were inpatients in psychiatric hospitals in Okayama Prefecture. (ii) Oral intake was difficult for them. (iii) Attending physicians judged that long-term artificial nutrition was necessary for survival. (iv) The decision on whether or not to make use of long-term artificial nutrition was made by attending physicians between January 1, 2014 and December 31, 2014. (v) Patients suffering from terminal cancer were excluded.
Artificial nutrition
Artificial hydration and nutrition includes enteral and intravenous nutrition. Enteral nutrition mainly consists of NG and PEG TF, while intravenous nutrition comprises peripheral venous nutrition (PVN) and total parenteral nutrition (TPN). TPN is usually used in the terminal state of malignancy and now rarely used for long-term care at psychiatric hospitals in Japan [10]. Patients receiving TPN were not evaluated in this study. The patients in both groups (TF or PVN) were fed orally before initiation of TF or PVN, and all of them had difficulty in eating orally during the 12 weeks before the decision.
Almost all inpatients in a terminal state in Japan receive artificial nutrition, and this study included no cases in a terminal state receiving both enteral nutrition and intravenous nutrition. Therefore, in this study, all patients not ingesting feeding tube nutrition received PVN in addition to oral intake.
Clinical diagnosis
All patients with Alzheimer’s disease (AD) were diagnosed according to the criteria for probable AD formulated by the NIA-AA [11]. All patients with vascular dementia (VaD) met the criteria for probable VaD of the AHA-ASA [12]. Other disorders were diagnosed according to ICD-10 criteria.
Questionnaires
Clinical characteristics of patients including age, sex, clinical diagnosis, methods of artificial nutrition, and duration of artificial nutrition were surveyed. Questionnaires on all subjects were completed by geriatric psychiatrists who knew the patient well and were chiefly in charge of the participants being evaluated. All raters had daily contact with the individuals being studied.
All medical records including nursing records and temperature tables were thoroughly examined by geriatric psychiatrists. They also evaluated the severity of dementia at the time of the decision whether or not to make use of long-term artificial nutrition by using the clinical dementia rating (CDR) [13] and functional assessment staging test (FAST) [14]. Physical comorbidity was evaluated using the Charlson Comorbidity Index (CCI) [15].
In patients receiving TF, records for a maximum 12 weeks before and 12 weeks after the start of TF were considered. The number of days of hospitalization in the psychiatric hospital was counted. The number of days when fever of 38 degrees and over was recorded, the number of days when intravenous antibiotics were used, and the number of bouts pneumonia during the observation period were counted.
Statistics
Statistical analyses were performed using IBM SPSS Statistics 23.0. Student’s t-test was used to compare two independent groups. Comparisons of proportions between two independent groups were calculated using a chi square test (2 × 2 table). The values for the same patient between before and after the intervention were compared using a paired t-test. The survival time of each group was plotted as a Kaplan-Meier survival curve, and survival times of groups were compared using a log-rank test. The effects of several variables (TF or PVN, age, sex, CCI scores) on survival time were investigated using Cox proportional hazards regression analysis. All p values were two-tailed, and p < 0.05 was accepted as significant.