Compliance of energy-dense, small volume oral nutritional supplements in the daily clinical practice on a geriatric ward — An observational study

Abstract

Introduction

Compliance is important in optimizing the clinical effectiveness of oral nutritional supplements (ONS). Small volume, energy-dense ONS (ED-ONS; ≥2 kcal/ml) have been shown to improve compliance in clinical trial settings. However, data from clinical practice is still lacking. The aim of the present study was to evaluate the effect of ED-ONS on the compliance in an observational set-up to obtain data from daily clinical practice on a geriatric ward.

Methods

Geriatric inpatients, undernourished or at risk of undernutrition received two servings of either ED-ONS (125ml, 2.4 kcal/ml: Nutridrink Compact Energy, Nutricia) or a standard ONS (S-ONS; 200ml, 1.5 kcal/ml: Nutridrink) as part of their daily routine care. Patients were allocated to a group according to availability of beds and placement on the ward. Compliance (kcal/day and % of prescribed volume) was assessed by weighing returned bottles. Data were analyzed via Mixed Model for Repeated Measures.

Results

Forty-seven patients received ED-ONS, and 61 patients received S-ONS. Compliance was significantly higher with ED-ONS in geriatric inpatients compared to S-ONS ( 378 ± 14.0 kcal/day vs. 337 ± 13.6 kcal/day (mean±SEM), p = 0.039, 63.0 ± 2.34% vs. 56.2 ± 2.26%, p = 0.039). Moreover, a trend (p=0.078) was observed towards an increasing difference in compliance over time.

Conclusion

This study shows that compliance to ED-ONS is significantly better than to S-ONS in daily clinical practice. Although small, the difference in compliance seems to increase over time, suggesting clinical relevance with longer treatment.

Introduction

Undernutrition is a common problem in the elderly, especially in acutely ill geriatric patients (1). Moreover, the severity of undernutrition generally increases during hospitalization (2). Nutritional intervention with ONS has been proven to be clinically effective in counteracting the detrimental effects of undernutrition 3, 4, 5. However, the effectiveness of ONS is strongly related to good compliance 5, 6.

Reports on compliance to ONS therapy in clinical trial settings range from poor (14.9%) (6) to excellent (97%) (7). In a systematic review, the overall mean compliance to all types of ONS regimens in all settings was 78% with a range of 37% to 100% (8). The compliance of hospitalized patients was lower than that of community dwelling elderly, namely 67% vs. 81% (8). Compliance to ONS is influenced by various factors including ONS related factors like flavours or texture variety 8, 9 and patient related factors, for example, assistance required in drinking ONS (10), dysphagia and co-morbidity. Many of the patient related factors are common in acutely ill geriatric patients. In addition, a negative correlation has been found between age and compliance to ONS (8).

The systematic review included a sub-analysis to compare the compliance of ONS with an energy density of ≥ 2 kcal/ml with compliance of ONS with energy density of 1 to 1.3 kcal/ml and 1.5 kcal/ml respectively, for patients of all ages in all settings (8). A significant positive correlation was demonstrated for energy density and compliance to ONS in clinical trial settings (8). Data on compliance to ONS of different volumes in daily clinical practice are lacking, therefore the aim of this study was to compare compliance of an energy dense, small volume ONS (ED-ONS; 300kcal/125ml= 2.4kcal/ml) with that of a standard ONS with the same energy content but a higher volume (S-ONS; 300kcal/200ml=1.5 kcal/ml) in geriatric inpatients with (a risk of) undernutrition in daily clinical practice. An observational set-up was chosen to evaluate compliance in daily clinical practice.

Material and method

Participants

From November 2011 to June 2012, all patients admitted to the geriatric ward of a large teaching hospital were screened for enrolment. To eliminate study-related influences, patients were not informed about the study during the observational period. Written informed consent to use the data was obtained from patients or their legal representatives after the observation period, just before discharge (or after the predefined maximal observation period of 6 weeks). Furthermore, the nursing staff was not informed about the exact objective of the study, and the clinical study did not interfere with normal clinical practice, including the departmental nutritional protocol.

Patients were included in the study if they were 65 years of age or older and had (a risk of) undernutrition according to the Mini Nutritional Assessment (MNA).

Exclusion criteria were an expected hospital stay of less than eight days, terminal illness, expected death during hospitalization, palliation, and need for tube or parenteral feeding. Patients who were included at admission and subsequently, due to unforeseen circumstances discharged before eight days were included in the analyses. The study protocol was approved by the local ethics committee.

Setting

The study was conducted on a 24 bed geriatric ward. On this ward, a multidisciplinary team is responsible for diagnosing and treating patients with geriatric syndromes.

The geriatric ward consists of two connected corridors with twelve patients residing on each corridor. Patient care is similar on both corridors, as one medical and nursing team cares for patients on both corridors. Only the colour of the walls of the corridors differ, i.e. one has red walls, the other yellow. Patients are admitted to the ward and thus the corridors in order of succession. No other criteria applied. Meals are served in a communal dining room, where assistance is available. Patients who are not bedridden are encouraged to keep a normal daily routine. Approximately half of the admissions are of an acute nature, for example delirium or a fall. Most patients are referred by their general practitioner.

Patients with (a risk of) malnutrition were prescribed two bottles of ONS per day according to the interdisciplinary nutrition protocol of the geriatric ward. According to this protocol every patient admitted to the ward is screened for undernutrition within 24 hours of admission, using the MNA. A score of less than 18 out of 30 indicates undernutrition, a score of 18 to 23.5 indicates a risk of undernutrition and a score of 24 or more means the patient is well nourished (11). ONS is served between three enriched meals and three snacks offered to all patients.

Intervention

Patients were assigned to either the red or the yellow corridor according to bed availability. Patients admitted to the red corridor were allocated to receive 2 bottles of S-ONS per day (200 ml of 1.5kcal/ml = 300kcal/bottle, 12 g protein/bottle). Patients on the yellow corridor were allocated two bottles of ED-ONS (125 ml of 2.4 kcal/ml = 300kcal/bottle, 12 g protein/bottle). The ONS were served in the original packaging, with a straw. Patients had a choice of strawberry, chocolate or vanilla flavour. Each day the administered bottles were collected. Returned bottles were taken to a secluded location in the hospital and weighed to the nearest gram by using a single digital scale (CAS SW-1S plus) by the researcher who was not a member of the multidisciplinary team, to ensure that the exact outcome measure of the study stayed unknown to staff on the ward. According to the departmental nutrition protocol, the total energy and protein intake (including supplements) was recorded on day 3 for patients with (a risk of) undernutrition. During this study the same assessment was repeated at day 8. Data from patients for whom intake was recorded before start of ONS supplementation were excluded in the analysis of total energy intake. Data from the day after admission until the day that consent was obtained were included in analysis.

Outcomes

For each patient age, sex, admission reason, duration of hospital stay, body mass index (BMI) and pre-clinical use of ONS were recorded. The primary outcome was compliance to ED-ONS compared to that of S-ONS. Compliance was defined as the proportion of prescribed ONS that was consumed and was assessed by calculating the difference between the weight of the returned bottles and the standard weight of dispensed bottles, recalculated to kcal/day and percentage of prescribed volume. For all bottles for which weight was recorded as unknown (UK) or not done (ND), intake was set to zero kcal. Secondary outcomes were total energy and protein intake on day 3 and day 8.

Data analysis

To compare baseline characteristics, independent samples Student's t-test or non-parametric Mann-Whitney tests were used for continuous parameters. Categorical and dichotomous parameters were compared using Pearson Chi-square test, or Fisher's exact test. Intake per day was compared between study groups using a mixed model repeated measurements (MMRM). The primary approach modeled the mean intake over time with a first order polynomial curve (straight line) using an autoregressive type I variance-covariance matrix. The mixed model included treatment, time and bottle as fixed effects, and subject as random effect. A treatment by time interaction term was included in the model and was subsequently removed if this interaction was not significant. Information on total energy and protein intake was only available at two time points (day 3 and day 8), and these variables were therefore analyzed using a fixed occasion MMRM model where time points are represented by dummies and subject as random factor. The difference in amount of missing data between the two groups were analysed using a Non-parametric Mann-Whitney U test. All statistical analyses was performed at Nutricia Research, using a two-side significant level of 5%, using SAS (version SAS 9.2, Entreprise Guide 4.1)

Results

Participants

Out of 325 screened patients, 170 patients were eligible for participation in the study. Permission to use the data was not obtained in 62 patients, mainly due to the fact that patients were discharged without being asked permission to use collected data. As informed consent has not been given, reason for admission and/or not giving permission cannot be described. Of the 108 patients who gave informed consent, 47 were admitted to the yellow corridor, thus receiving the ED-ONS. Sixty-one patients were admitted to the red corridor receiving the S-ONS. Subject flow chart is described in figure 1.

Figure 1.

Patients flow chart

Estimated mean intake (kcal/day), calculated by a Mixed Model for Repeated Measurements (including treatment as fixed factor and subjects as random factor) [lines], and raw means [dots] combined; P=0.078 for interaction effect “treatment by time”.

The mean (±SD) age of the population was 83.1 ± 6.2 years, mean BMI was 25.5 ± 4.6. Both groups were comparable with respect to baseline characteristics (table 1). Compliance recalculated to kcal/day was 378 ± 14.0 kcal/day (mean±SEM) for the ED-ONS and 337 ± 13.6 kcal/day for the S-ONS, (p = 0.039). Recalculated to percentage compliance was 63.0 ± 2.34% for the ED-ONS and 56.2 ± 2.26% for the S-ONS. Moreover, a trend (p = 0.078) was observed towards an increasing difference in compliance over time (Fig 2). Age, height, body weight, BMI, MNA total score, duration of the observational period, sex, user-type and reason for admission was checked as possible confounders. None of these had a statistically significant effect on the outcome (table 1).

Table 1.

Baseline characteristics

	ED ONS group n = 47	Standard group n = 61	p
Sex female	n (%)	29 (61.7)	40 (65.6)	0.6781
Age (years)	mean ± SD	83.4 ± 5.4	82.9 ± 6.7	0.7022
BMI (kg/m2)	n	45	61	0.1622
	mean ± SD	26.2 ±4.35	25.0 ± 4.69
MNA	n	46	60	0.1553
	median (min-max)	18.5 (5.5-23.5)	19.5 (7.0-23.5)
Hospital stay (days)	n	47	59	0.0743
	median (min-max)	10 (2-28)	9 (1-25)
ONS use
At admission	n (%)	1 (2.1)	2 (3.3)
New	n (%)	46 (97.9)	59 (96.7)	1.0004
Admission reason
1. Cognitive/mood disorder	n	5	7	0.8911
2. Fall/mobility problem	n	3	11	0.0741
3. Pain	n	9	10	0.7091
4. Delirium	n	6	7	0.8381
5. Infection	n	4	4	0.7264
6. Functional de-cline	n	2	0	0.1874
7. Gastro-intestinal problems	n	5	14	0.0961
8. Anemia	n	5	3	0.2911
9. Other	n	8	5	0.1621

Mini Nutritional Assessment (MNA), Body mass index (BMI), 1. Chi-squaretest; 2. T-test for independent samples; 3. Non-parametric Mann-Whitney U test; 4.Fisher^vs exact test

Figure 2.

Compliance over time

No significant difference was observed on day 3 and day 8 in total energy and protein intake between both groups (table 2). However, there was a trend (p=0.087) for an increase in energy intake over time in the ED-ONS group. Bottles were not returned for a median of 10% of servings of ONS in the ED-ONS group, and 13% of servings in the S-ONS group. (p = 0.304)

Table 2.

Total energy and protein intake on day 3 and day 8

	ED-ONSgroup		Standard ONS group
	n	mean ± SEM1	n	mean ± SEM1	P1
Energy (kcal/day)
Day 3	32	1547 ± 82.1	39	1659 ±74.1	0.312
Day 8	31	1695 ± 65.4	39	1608 ± 58.2	0.323
Protein (g/day)
Day 3	32	70.1 ± 3.54	39	68.4 ± 3.24	0.677
Day 8	31	72.1 ± 3.29	39	70.4 ± 2.95	0.677

Data was not included in analysis for patients for whom intake was recorded before start of ONS, 1; Estimated marginal mean ± SEM and p-value from Mixed Model for Repeated Measurements, including treatment as fixed factor and subjects as random factor.

Discussion

Our hypothesis that energy density and small volume of ONS would influence the compliance in clinical practice with ill geriatric patients is confirmed in this study. The compliance to ED-ONS is higher than to S-ONS, supporting findings in previously published studies, in different settings and populations (8). Taking into account the relatively short duration of treatment and the trend towards increased difference in intake between the ED-ONS and the S-ONS over time, better compliance can be expected if ONS is used over a longer period. The average duration of hospitalization for patients in this study was 10.6 days, whereas according to the consensus guideline for treatment of undernutrition in Dutch geriatric patients, treatment of undernutrition should last three months (12).

Achieving the recommended energy and protein intake is a challenge in hospitalized patients and food wastage is high (13). Better compliance to the ED-ONS in our study leads to a higher ONS related energy intake compared to standard ONS, accounting for about 2.5% of the total daily intake. Although the observed difference is low, this percentage is thought to increase with longer follow-up, according to the trends observed in our study. Factors related to the acute illness of the patients negatively influence nutritional intake to the degree that every additional intake of energy and protein counts. Compliance to ONS has been found to positively correlate to patients total energy intake (8). Although total energy intake increased in the ED-ONS group but not in the standard ONS group, no significant difference in energy intake was observed between both groups in the present study. This might be explained by the large variation in dietary intake between the groups.

A strong aspect of the present study was that compliance was measured in an objective manner by weighing the difference between issued and consumed ONS, thereby eliminating bias caused by patient recall. The study was designed so that patients were unaware of the fact that compliance was being measured until informed consent was asked. The nursing staff was not informed that the goal of the study was to measure compliance, to eliminate the effect of extra encouragement from their part to drink ONS. These measures ensure an unbiased reflection of compliance to ONS in a hospital setting. To our knowledge a difference in compliance to ONS of different volumes has not yet been proven in such an objective manner without interfering with daily practice.

The strength of this study, namely its design to obtain data from a real life setting without influencing both patient and personnel, unfortunately also leads to one of its weaker points, which is missing data. As we could only ask for informed consent at discharge, data from a considerable number of patients could not be used as it was not feasible to give information and ask for permission during the sometimes complex discharge process.

Furthermore, although nutrition has a high priority on the ward, the reality is that in situations with acutely ill patients, nutrition receives less attention than it probably should. Other components of treatment may seem more important during the acute phase. As a result, a median of 10 % of servings in the ED-ONS group and 13% of servings in the standard group that should have been dispensed and returned, were missing. Nevertheless, the amount of missing data is still lower compared to what we observed in outpatients’ diaries in randomized clinical trials. Missing data in the present study can be explained by two possible scenarios. Bottles that were dispensed could have been disposed of by accident. Or the second serving of ONS was not given if the first one was not empty yet.

In conclusion this study shows that a lower volume and higher energy density ONS leads to higher rates of compliance in acutely ill geriatric inpatients. Thus, prescription of ED-ONS instead of normal volume ONS could have a better effect in the treatment of under-nutrition in acutely ill geriatric inpatients. Moreover, with the observed trend towards increased difference in compliance over time, it could benefit patients treated for 3 months as recommended. More studies are needed to prove a direct relation between clinical improvement and the use of ED-ONS as opposed to normal volume ONS, specifically focused on longer term interventions than the current study.