Effect of selective gastric residual monitoring on enteral intake in preterm infants

Gastric residuals (GRs) evaluation consists in measuring the volume of milk along with gastrointestinal secretions remaining in the stomach after a certain time interval [1]. It is common to find GRs in preterm infants due to the immaturity of the gastrointestinal tract in the terms of decreased length, immature motility patterns, and inadequate digestive and absorptive capacity [2, 3]. It has been postulated that the negative pressure created by repeated aspiration of gastric contents might damage the fragile gastric mucosa, and discarding the gastric contents results in a loss of gastric enzymes and gastric acid in preterm infants. However, the use of GRs as a marker of feeding intolerance [4] is widespread in neonatal intensive care units (NICUs) and has contributed to making a diagnosis of feeding intolerance very frequent. It has been estimated that 16 to 29% of premature infants admitted to NICUs develop a feeding intolerance [5, 6] whose subsequent management depends on nurses’ experience, clinicians’ individual preference, and NICU protocols. On the other hand, feeding intolerance has been associated with an increased risk of developing necrotizing enterocolitis (NEC) [7, 8] and, therefore, an increased volume of altered GRs (bilious- or blood-stained) has often, in turn, been considered a risk factor for the development of NEC. Thus, the importance and diffusion of GRs (and feeding intolerance) monitoring depends on its possible usefulness in the prevention or early diagnosis of NEC, which remains the most frequent neonatal gastrointestinal complication in preterm infants [9, 10] increasing their mortality ⁸ and risk of neurodevelopment delay [11]. Nevertheless, the role of prefeed GRs monitoring to detect feeding intolerance (and a potential risk of NEC) is controversial and has been correlated to an overestimation of feeding intolerance increasing the number of feed interruptions and the time taken to reach full feeds without beneficial effect on the occurrence of NEC [1, 12, 13]. These results are relevant because achievement of full enteral feeding allows better preterm infant nutrition and growth, prompt discontinuation of parenteral nutrition and reduction of catheter-related bloodstream infections [1, 14]. Based on these considerations, we decided to implement a new local protocol in our NICU that provides for selective rather than routine evaluation of GRs with the assumption that this might allow earlier full enteral feeding to be achieved and improvement of patients’ growth. Thus, the aim of this study was to evaluate the effects of the new strategy on these outcomes in comparison to the previous approach.

A total of 118 infants were enrolled in the study, 49 in the selective GR group and 59 in the routine GR group. Clinical characteristics, including prematurity complications and particularly the occurrence of NEC (6 vs. 2%; P = 0.320), were similar between the groups (Tables 1 and 2).

Age at full (150 mL/kg) enteral feeding (17.8 ± 10.1 vs. 22.9 ± 10.5 days, P = 0.017), full oral enteral feeding (56.2 ± 6.1 vs. 58.3 ± 4.0 days, P = 0.045), and birth weight recovery (11.1 ± 3.0 vs. 12.5 ± 3.5 days, P = 0.039) was lower while Z-scores of weight at discharge (-1.10 ± 0.83 vs. -1.60 ± 1.45, P = 0.040) were higher in infants in the selective GR group in comparison to infants in the routine GR group (Tables 2 and 3). We observed a trend toward shorter duration of parenteral nutrition, stay in NICU and in hospital in the no routine residual group in comparison with controls, but differences did not reach statistical significance (Table 2). Exclusive and mixed breastmilk feeding at discharge had similar occurrence in the groups (Table 2).

In this study we evaluated the effects of implementing in our NICU a new protocol for the monitoring of GRs through a selective monitoring approach rather than routine monitoring. We found that the new procedure allowed achievement of full enteral feeding and full oral (breast and bottle) enteral feeding sooner. These results are in agreement with previous studies. Abiramalatha et al. meta-analysed two randomised controlled studies which enrolled 141 very low birth weight infants and found that routine monitoring of GRs increases the time needed to establish full enteral feeds, to regain birth weight, and the s duration of parenteral nutrition [1]. Kumar et al. meta-analysed six studies which enrolled 451 preterm infants and found that avoidance of routine prefeed evaluation of GRs was associated with earlier achievement of full enteral feeding, shorter duration of hospitalization, and lower incidence of late-onset sepsis [20]. These results are explained by the inappropriate discontinuation of enteral feeding with subsequent delays in advancement of enteral nutrition associated with routine prefeed assessment of GRs. In fact, it is well known that GR evaluation is not an accurate marker of feeding intolerance due to a lack of standardization regarding evaluation of gastric tube placement, also in relation to patient position [20], the most appropriate timing of assessment, the volume and/or color that should be considered abnormal and, moreover, the use of GRs which could be returned or discarded [21].

Our study confirms that avoiding routine monitoring of GRs is not associated with an increase of incidence of NEC. A similar result was reported by several studies [12, 13, 21, 22] and meta-analyses [1, 23] and, more recently, by Purohit et al. who found a lack of association between GRs and the risk of developing a NEC ≥ stage II [24]. These findings are consistent with the study by Mihatsch et al. who concluded that assessment of GRs (volume and color) is not only not predictive of feeding intolerance, but also not predictive of NEC [25]. Other studies [26, 27] have reported, in contrast, an association between increased GRs and NEC, but they were retrospective and had many confounding factors which limit their value. Thus, GRs alone are an unreliable marker of NEC, while in association with other suspicious clinical signs [16] they can contribute to a rational management of enteral nutrition and to the planning of any diagnostic tests.

We found that the time to regain birth weight and Z-scores of weight at discharge were significantly better in infants in the no routine residual group than in the control group, while Z-scores of length and head circumference were similar. These findings are in agreement with previous studies [1, 12, 13] and confirm the positive effect on nutrition and growth of a selective GRs evaluation policy which accelerates the time needed to reach full enteral feeding and improves the nutritional intake of these infants.

The main limitation of this study is that it was a retrospective single center study based on a “before and after” design which cannot rule out the possibility of unmeasured confounding factors contributing to the results. However, during the study periods clinical practice at the center did not change and our findings were substantially in agreement with previous studies. These considerations, together with the similar demographic and clinical characteristics of the study groups, support the reliability of our results. Another limitation regards the size of our population which did not allow us to find statistical significance in the shortening of parental nutrition, stay in NICU and in hospital duration that we observed as a trend in the selective GR monitoring group in comparison with the routine GR monitoring group. Furthermore, although we have not observed significant changes in the onset of NEC, our study does not have sufficient power to definitively exclude the risk of its increase. However, despite these limitations, we believe our study adds useful contemporary data about implementing a new protocol of monitoring in term of its effects and safety in a NICU.

In conclusion, we found that the selective monitoring of GRs in a cohort of extremely preterm infants was associated with a decrease of age at full enteral feeding and at birth weight recovery and with better Z-scores of weight at discharge in comparison with routine GR monitoring. The incidence of NEC did not increase although this finding, while in agreement with previous studies, should be confirmed in a larger population. The assessment of GRs is time consuming and devoid of beneficial effects. Thus, its current widespread application seems no longer justified according to literature evidence, while it seems reasonable to omit prefeed GRs monitoring in clinical practice.