Original Article 19
Nasogastric vs. Intravenous Rehydration in Children with Gastroenteritis and Refusal to Drink: A Randomized Controlled Trial
Authors
J. Marquard1, C. Lerch2, A. Rosen1, H. Wieczorek1, E. Mayatepek1, T. Meißner1
Affiliations
1
Department of General Paediatrics, Neonatology and Paediatric Cardiology, University Children’s Hospital Duesseldorf, Germany 2 Department of General Practice, Cochrane Metabolic and Endocrine Disorders Group, Medical Faculty, University Hospital Duesseldorf, Germany
Key words ▶ gastroenteritis ● ▶ moderate dehydration ● ▶ nasogastric rehydration ● therapy ▶ intravenous rehydration ● therapy ▶ children ●
Abstract
Zusammenfassung
Background: Nasogastric rehydration therapy (NGRT) is the recommended therapy in moderately dehydrated children with gastroenteritis and refusal to drink, since it is supposed to be as effective if not better than intravenous rehydration therapy (IVRT). However, in clinical practice IVRT is often favored. We conducted a clinical trial to determine whether IVRT is not inferior to NGRT. Patients and Methods: Children 3 months to 6 years of age with moderate dehydration and refusal to drink secondary to gastroenteritis were recruited. After clinical assessment of the degree of dehydration, patients were assigned randomly to receive either IVRT or NGRT over 6 h on the hospital ward. Results: Recruitment did not yield the estimated number of patients. Mainly, non-enrollment was due to failure to obtain parental consent because IVRT was expected. 97 patients were enrolled in the study, 46 were randomized to NGRT and 51 to IVRT. There was no difference between IVRT and NGRT groups concerning length of hospital stay (2.2 ± 1.1 days vs. 2.4 ± 1.1 days), success of rehydration (78 vs. 76 %) and adverse events. Discussion: Since we had to terminate the study ahead of schedule due to a low recruiting rate, our results are not reliable. However, data from the literature shows that the widespread described superiority of NGRT over IVRT is seriously influenced by studies from developing countries questioning the applicability of the results to a setting available in high-income countries nowadays. Conclusion: Our study demonstrates the difficulties performing such a study in a high-income country to come to an objective and clearly evident final conclusion.
Hintergrund: Bei Kindern mit moderater Dehydratation, akuter Gastroenteritis (aGE) und Nahrungsverweigerung wird eine nasogastrale Rehydratationstherapie (NGRT) empfohlen. In der klinischen Praxis wird jedoch häufig eine intravenöse Rehydratationstherapie favorisiert (IVRT). Wir untersuchten im Rahmen einer klinischen Studie, ob eine NGRT der IVRT unterlegen ist. Patienten und Methode: Eingeschlossen wurden Patienten im Alter von 3 Monaten bis 6 Jahre mit moderater Dehydratation, aGE und Nahrungsverweigerung. Nach klinischer Einschätzung des Schweregrades der Dehydratation wurden die Patienten randomisiert dem Behandlungsarm NGRT oder IVRT zugeteilt und über 6 h unter stationären Bedingungen rehydriert. Ergebnisse: Die geplante Fallzahl wurde nicht erreicht, viele Eltern lehnten die Studienteilnahme ab, da sie eine IVRT erwarteten. Insgesamt wurde 97 Patienten in die Studie eingeschlossen (46 NGRT, 51 IVRT). Es zeigten sich keine Unterschiede in der Länge des Krankenhausaufenthaltes (2,2 ± 1,1 Tage IVRT vs. 2,4 ± 1,1 Tage NGRT), beim Erfolg der Rehydratationstherapie (78 % IVRT vs. 76 % NGRT) sowie unerwünschten Ereignissen. Diskussion: Da die Studie vorzeitig beendet wurde, können keine Schlussfolgerungen aus den Ergebnissen gezogen werden. Dennoch ist zu erwähnen, dass die häufig beschriebenen Vorteile der NGRT über die IVRT auf Studien aus Entwicklungsländern beruhen und sich nicht auf die heutige Therapie in Industriestaaten übertragen lassen. Schlussfolgerung: Unsere Studie zeigt die Schwierigkeiten der Durchführbarkeit einer Studie in diesem Bereich, um zu klaren Ergebnisse zu kommen.
Schlüsselwörter ▶ Gastroenteritis ● ▶ moderate Dehydratation ● ▶ nasogastrale Rehydratation ● ▶ intravenöse Rehydratation ● ▶ Kinder ●
Bibliography DOI http://dx.doi.org/ 10.1055/s-0033-1363245 Published online: January 16, 2014 Klin Padiatr 2014; 226: 19–23 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0300-8630 Correspondence Dr. Jan Marquard Department of General Paediatrics Neonatology and Paediatric Cardiology University Children’s Hospital Duesseldorf Moorenstraße 5 40225 Düsseldorf Germany Tel.: + 49/211/8117 711 Fax: + 49/211/8119 512 [email protected]
▼
▼
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
Nasogastrale vs. intravenöse Rehydratation bei Kindern mit Gastroenteritis und Nahrungsverweigerung: Eine randomisierte kontrollierte Studie
Background
▼
According to the guidelines of the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition [6] and the American Academy of Pediatrics [8] oral rehydration therapy (ORT) should be used as first-line therapy for the management of gastroenteritis in children. When ORT by mouth fails, nasogastric rehydration therapy (NGRT) should be performed, since it is supposed to be as effective if not better than intravenous rehydration therapy (IVRT) [6]. However, in this situation IVRT is often favored in clinical practice: 75 % of American emergency medicine physicians who indicated familiarity with the guidelines gave IVRT to patients with moderate dehydration [11]. Recently, data from a Canadian pediatric emergency department have been published showing that if oral rehydration fails, only 10 % of caregivers and 14 % of clinicians would choose NGRT [2]. Moreover, caregivers were more familiar with the term intravenous than nasogastric rehydration [2]. This might also be the case for institutions in Europe. Another point is that blood specimen is usually drawn in dehydrated children. Blood withdrawal can be performed with the same procedure the intravenous line is placed. However, when NGRT is the therapy of choice, additional nasogastric tube insertion beside blood withdrawal is always more annoying for the patient. Interestingly, the guidelines for rehydration therapy in children are partially based on studies from low-income countries published in the 1980s or on studies with very small sample sizes. Since we doubt that the studies cited in the guidelines are generally representative for the treatment of moderately dehydrated children with acute gastroenteritis and refusal to drink in high-income countries nowadays, we designed a randomized controlled trial to evaluate the hypothesis that IVRT would not be worse than NGRT. The primary objective of our study was to determine the length of hospital stay (LOS) of moderately dehydrated children rehydrated by nasogastric tube vs. peripheral intravenous catheter (PVC); secondary objectives were to determine the success of rehydration therapy and the occurrence of adverse events.
Methods
▼
Patients and trial design Patients were recruited between August 2008 and December 2011, in the emergency department (ED) of a Children’s University Hospital in Germany. The local Ethics Committee of the Medical Faculty approved the study (approval number 3075). Written informed consent was obtained from all parents. During the study period, all children with gastroenteritis who presented in the emergency department were screened for eligibility. Inclusion criteria were: 1. Age 3 months to 6 years; 2. Acute diarrhea defined as ≥ 3 loose stools of sudden onset in previous 24 h for ≤ 7 days, with or without fever or vomiting [8]; 3. Diagnosis of moderate dehydration using the dehydration score system described by Gorelick et al. (dehydration score ≥ 3 and < 7, corresponding to 5–10 % dehydration) [4] and 4. Refusal to drink, defined as inability of fluid intake tested in the emergency department. The dehydration score by Gorelick includes the parameter “Decreased skin elasticity”, “Capillary refill > 2 s”, “General appearance”, “Absent tears”, “Abnormal respiration”, “Dry mucous membranes”, “Sunken eyes”, “Abnormal radial pulse”, “Tachycardia (heart rate > 150 beats per min)” and “Decreased urine output” [4].
Exclusion criteria were significant other morbidity, pre-treatment in any emergency department or hospital in the preceding 24 h and any injury or abnormality in the nasopharyngeal space. Patients were randomized, as described in detail below, to receive either IVRT or NGRT. For children randomized to IVRT, an intravenous line was placed by a physician, in the NGRT group a nasogastric tube was inserted by a nurse. To make sure that the nasogastric tube is at the right place fluid drawn out from the tube was tested for acidity with a pH indicator to show whether or not the tube is in the stomach. A maximum of 3 attempts at intravenous line or nasogastric tube insertion was allowed. If more than 3 attempts failed, physicians respectively nurses were recommended to cross over to the other treatment. Blood was collected in both groups (intravenous blood collection in the IVRT, capillary blood collection in the NGRT group) to perform a blood gas analysis including blood electrolytes and blood glucose. For rehydration therapy, the patients were moved to the ward. A pre-therapy naked weight in children with a body weight less than 10 kg was measured to the nearest 0.01 kg using a SECA electronic baby scale by nursing staff, in children with a body weight more than 10 kg body weight in light indoor clothes was measured to the nearest 0.1 kg using a SECA chair scale [SECA, Hamburg, Germany]. Patients were rehydrated during a 6-h rehydration phase: in the IVRT group they received 50 ml/kg of a 2.5 % glucose in 0.45 % sodium chloride solution over 6 h via an infusion pump. In the NGRT group they received the same amount of oral rehydration solution containing 60 mmol/L sodium, 20 mmol/L potassium, 60 mmol/L chloride, 10 mmol/L citrate and 90 mmol/L glucose over 6 h at a constant rate through a nasogastric tube. In our hospital, the use of oral ondansetron is uncommon in children with gastroenteritis and was not used in the study. The local Ethics Committee recommended the use of racecadotril in all study patients because it was a standard treatment for children with gastroenteritis with 3 or more watery or loose stools within 24-h in our hospital. The dose was 1.5 mg/kg, 3 times daily at regular intervals until 2 normal stools were recorded. In the NGRT group racecadotril was administered via the nasogastric tube, in the IVRT group children had to drink racecadotril, added to water. Since acute diarrhea defined as ≥ 3 loose stools of sudden onset in previous 24 h was an inclusion criterion, all patients received racecadotril. However, since refusal to drink was an inclusion criterion as well, the intake of racecadotril was more common in the NGRT group. This has seriously influenced the study in favor for NGRT. After 6 h of rehydration, patients were reassessed according to the dehydration score system [8] for signs of dehydration, and body weight was measured again, under the same conditions as prior to treatment. If refusal to drink was ongoing and or rehydration was considered as “not successful”, maintenance fluid requirement was administered according to the formula: First 10 kg 100 ml/kg/day, second 10 kg 50 ml/kg/day and subsequent kg 20 ml/kg/day. If necessary, physicians adjusted the rate of rehydration to replace the volume of ongoing losses. During the hospitalization, oral fluid intake, number of vomiting, and diarrhea were measured constantly. Patients were discharged once oral feeding was tolerated (defined as oral fluid intake > 25 % of daily recommended fluid amount in last 6 h), in the absence of clinical signs of dehydration and vomiting, and less than 5 loose stools in the last 24 h. A follow-up visit was not performed within this study.
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
20 Original Article
Original Article 21
Fig. 1 Flow diagram. Assessed for eligibility (n= 4 715)
Excluded (n= 4618) ◆ Not meeting inclusion criteria (n= 4439) ◆ Declined to participate (n= 179)
Allocated to NGRT (n= 46) ◆ Received allocated intervention (n= 43) ◆ Did not receive allocated intervention (n= 3) (unable to obtain nasogastric intubation)
Allocated to IVRT (n= 51) ◆ Received allocated intervention (n= 51) ◆ Did not receive allocated intervention (n= 0)
◆ Lost to follow-up (n= 0) ◆ Discontinued intervention (n= 12) (persistent vomiting)
◆ Lost to follow-up (n= 0) ◆ Discontinued intervention (n= 0)
46 analyzed (intention-to-treat)
51 analyzed (intention-to-treat)
Outcome measurements The primary outcome was LOS in days, calculated by subtracting day of admission from day of discharge (equivalent to number of nights). Minimum LOS was 1 day (1 night). LOS in hours was not calculated since discharge of patients was performed within the normal routine during and day and therefore could have biased the results. Secondary outcomes included success of rehydration therapy, defined as resolution of moderate dehydration (a 6-h dehydration score ≤ 2, weight gain and production of urine (defined as urine in nappy, urine volume was not determined) within the rehydration phase). NGRT was changed to IVRT when treatment was not tolerated, e. g. due to persistent vomiting (defined as ≥ 3 episodes of vomiting). Occurrence of adverse events was defined as diagnosis of therapy related events like phlebitis or extravasation injuries in the IVRT group. Body temperature, vital signs, urine volume and number of placement failures in the IVRT or NGRT group were not analyzed in this study. Parameters of pain and dissatisfaction of the patients were not documented. Cost effectiveness of both procedures was not considered as outcome measure.
sequence that was stratified by age ( ≤ 3 years vs. > 3 years) and applied 3 different random block sizes. Sequentially numbered, opaque and sealed envelopes concealed the allocation sequence from the clinicians.
Data analyses Data was recorded on a case record form. All analyses followed the intention to treat principle and included patients with protocol deviations. Analyses were performed with SPSS software (version 11.5.1, 2002). Data was presented as means ± standard deviation (SD), medians with interquartile range (IQR), absolute numbers or specified in percent. Categorical data was analyzed by Chi-Square test, continuous data was analyzed by MannWhitney test. For the primary outcome parameter LOS we also calculated the difference in means and its 95 % confidence interval using a 2-sample t-test. A p-value of < 0.05 was considered as statistically significant.
Results
▼
Patients Sample-size calculation Data from our hospital indicated that the standard deviation of LOS was 1.5 days. To detect a reduction in LOS of 0.6 days (that is, 3 out of 5 children would be spared one night in hospital) with a 2-sided α of 5 %, a power of 80 % and an anticipated dropout rate of 10 %, a sample size of 110 patients per group was necessary [1].
Randomization Allocation of the participants occurred according to a computer generated randomization list. Stata 10 (College Station, Texas, USA: StataCorp LP) was used for creating the 1:1 randomization
During the study period, 4 715 patients presented to the emergency department with clinical signs of acute gastroenteritis. 276 eligible patients were identified of whom 97 (35.1 %) were enrolled in the study, 46 patients were allocated to NGT and 51 to IVRT. Caregivers of 179 patients (64.9 %) declined to participate. In 3 patients of the NGT group, placement of nasogastric tube ▶ Fig. 1). Baseline characfailed (6.5 %) and IVRT was performed (● ▶ Table 1). teristics were similar between the study groups (● The recruitment did not yield the number of patients estimated in the sample size calculation. Since we had the impression that there was an increased rate of change from NGRT to IVRT due to vomiting, we decided to conduct a non pre-specified interim
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
Randomized (n= 97)
22 Original Article
analysis concerning ‘Allocated intervention not tolerated’ after 97 patients were included. The analysis showed that change from NGRT to IVRT was significant and we decided to end ahead of schedule. Another critical point was a recruitment rate lower than anticipated. Reason for non-enrollment of eligible patients was failure to obtain parental consent. Only the intention to treat analysis is reported since the results were similar to those of the per-protocol analysis.
ing rehydration (clinical diagnosis made by the ward physician) and 8 did not tolerate the therapy due to persistent vomiting (17 %). The overall rate of change of rehydration therapy was significantly more common in the NGRT group compared to the IVRT group (26 vs. 0 %, p < 0.001). One patient in the IVRT group had a mild extravasation injury, there was no patients with phle▶ Table 2). In the NGRT group no adverse events were bitis (● reported during the treatment phase (e. g. misplacement of tube, removal by the child).
Primary outcome
Secondary outcomes ▶ Table 2 gives details about secondary outcomes. No difference ● between the 2 groups in overall success of rehydration at 6 h could be found (78 % in the IVRT group and 76 % in the NGRT group). Resolution of moderate dehydration and weight gain was found to be the same in both groups. However, production of urine was statistically more common in the IVRT group compared to the NGRT group (98 % vv. 90 %, p = 0.035). Of the 46 patients in the NGRT group, 4 (9 %) had clinical impairment durTable 1 Baseline characteristics of the study sample. Variable age (years) mean median sex ( % female) baseline dehydration score mean median blood base excess (mmol/l) median blood ph median blood sodium (mmol/l) median blood potassium (mmol/l) median leucocytes (x109/l) median
IVRT (n = 51)
NGRT (n = 46)
1.7 ± 1.1 1.3 (0.8–2.3) 41 (n = 21)
1.8 ± 1.1 1.4 (1.0–2.3) 43 (n = 20)
3.6 ± 0.7 3.0 (3.0–4.0)
3.6 ± 0.7 3.5 (3.0–4.0)
− 7.3 ( − 9.8 to − 5.1)
− 8.1 ( − 9.3 to − 5.1)
7.37 (7.36–7.40)
7.38 (7.36–7.40)
136 (135–138)
138 (136–139)
3.9 (3.6–4.3)
4.2 (3.8–4.5)
8.6 (5.9–11.9)
7.8 (7.1–10)
Data are shown for the intention-to-treat population. Data are given as median with interquartile range or specified in percent. IVRT intravenous rehydration therapy, NGRT nasogastric rehydration therapy
Variable length of hospital stay (days) mean median successful rehydration ( %) resolution of moderate dehydration (score ≤ 2) ( %) production of urine ( %) weight gain ( %) change of rehydration therapy ( %) clinical impairment during rehydration ( %) allocated intervention not tolerated ( %) complications phlebitis mild extravasation injury ( %)
Discussion
▼
After termination the study ahead of schedule, results of the interim analysis show no difference in LOS, success of rehydration therapy, and occurrence of adverse events between IVRT and NGRT in moderately dehydrated children with acute gastroenteritis and refusal to drink in the setting of our hospital. However, it has to be mentioned that we cannot draw any conclusions out of these data since the recruitment did not yield the number of patients estimated in the sample size calculation. Interestingly, the recruitment rate was dramatically lower than anticipated. Caregivers of eligible patients were informed on both rehydration regimes. If they declined to participate, they could still choose between the 2 options (NGRT or IVRT). Remarkably, all non-enrolled eligible patients were rehydrated intravenously, suggesting parental preference for IVRT. Reason for non-enrollment of eligible patients was failure to obtain parental consent because caregivers expected IVRT, were familiar with IVRT or believed that NGRT is more painful. However, IVRT was the established rehydration therapy in children with gastroenteritis in our hospital prior to this study, so the physicians’ tendency to IVRT could also have influenced parents in their decision on therapy regime or on willingness to participate in this study. This finding is similar to Freedmans et al. [2] results showing that caregivers are more familiar with the term IVRT and believe that intravenous insertion is easier compared to nasogastric insertion. Despite termination the study ahead of schedule, there is one important finding: Parental and physician’s preference for IVRT seems to be tremendously high in Germany and probably in other high-income countries as well. Therefore the feasibility of our study design is called into question. However, we still see a rationale for the question whether IVRT should also be a recommended therapy in high income countries since the guidelines for rehydration therapy in children are
IVRT (n = 51)
NGRT (n = 46)
p-Value
2.2 ± 1.2 2 (1–3) 78 (40/51) 92 (47/51) 98 (50/51) 96 (49/51) 0 (0/51) 0 (0/51) 0 (0/51)
2.4 ± 1.2 2 (1.3–3) 76 (35/46) 90 (40/46) 90 (40/46) 91 (42/46) 26 (12/46) 9 (4/46) 17 (8/46)
0.34 0.78 0.4 0.035 0.33 < 0.001 0.031 0.002
0 (0/51) 2 (1/51)
0 (0/46) 0 (0/46)
Data are given as means ± SD, median with interquartile range or specified in percent
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
0.34
Table 2 Primary and secondary outcome measurements.
Downloaded by: National University of Singapore. Copyrighted material.
There was no difference between the 2 groups in terms of LOS: 2.2 ± 1.1 days in the IVRT group and 2.4 ± 1.1 days in the NGRT ▶ Table 2) (difference in LOS − 0.2 days; 95 % confidence group (● interval, − 0.689–0.3036; p = 0.44).
Original Article 23 Conclusion
▼
There is no doubt that ORT by mouth is the treatment of choice in dehydrated children with gastroenteritis. If children are able to drink oral rehydration solution they do not require a nasogastric tube or peripheral venous catheter, both potentially painful and difficult procedures. When ORT by mouth fails, NGRT is the recommended therapy in Germany. Although rehydration is a frequent daily procedure in German pediatric hospitals, there are no reliable data for Germany or high-income countries available showing that NGRT is superior to IVRT for rehydration of moderately dehydrated children with gastroenteritis and refusal to drink. Since we had to terminate the study ahead of schedule due to a low recruiting rate, we are not able to answer this question reliable based our results. However, parental and physician’s preference for IVRT seems to be tremendously high in Germany and probably in other high-income countries as well. Our study demonstrates the difficulties performing such a study in a high-income country to come to an objective and clearly evident final conclusion.
Conflict of interest: The authors have no conflict of interest to disclose. References 1 Dupont WD, Plummer WD. Power and Sample Size Calculations: A Review and Computer Program. Control Clin Trials 1990; 11: 116–128 2 Freedman SB, Keating LE, Rumatir M et al. Health Care Provider and Caregiver Preferences Regarding Nasogastric and Intravenous Rehydration. Pediatrics 2012; 130: e1504–e1511 3 Gonzalez-Adriano SR, Valdes-Garza HE, Garcia-Valdes LC. Oral hydration versus intravenous hydration in patients with acute diarrhea. Bol Med Hosp Infant Mex 1988; 45: 165–172 4 Gorelick MH, Shaw KN, Murphy KO. Validity and reliability of clinical signs in the diagnosis of dehydration in children. Pediatrics 1997; 99: E6 5 Gremse DA. Effectiveness of nasogastric rehydration in hospatalized children with acute diarrhea. JPGN 1995; 21: 145–148 6 Guarino A, Albano F, Ashkenazi S et al. European Society for Paediatric Gastroenterology, Hepatology, and Nutrition/European Society for Paediatric Infectious Diseases: Evidence-based Guidelines for the Management of Acute Gastroenteritis in Children. JPGN 2008; 46: S81–S122 7 Hartling L, Bellemare S, Wiebe N et al. Oral versus intravenous rehydration for treating dehydration due to gastroenteritis in children. Cochrane Database Syst Rev 2006; 3: CD004390 8 King CK, Glass R, Bresee JS et al. Centers for Disease Control and Prevention: Managing acute gastroenteritis among children: oral rehydration, maintenance, and nutritional therapy. MMWR Recomm Rep 2003; 21: 1–16 9 Mackenzie A, Barnes G. Randomised controlled trial comparing oral and intravenous rehydration therapy in children with diarrhoea. BMJ 1991; 303: 393–396 10 Nager AL, Wang VJ. Comparison of nasogastric and intravenous methods of rehydration in pediatric patients with acute dehydration. Pediatrics 2002; 109: 566–572 11 Ozuah PO, Avner JR, Stein RE. Oral rehydration, emergency physicians, and practice parameters: a national survey. Pediatrics 2002; 109: 259–261 12 Sharifi J, Ghavami F, Nowrouzi Z et al. Oral versus intravenous rehydration therapy in severe gastroenteritis. Arch Dis Child 1985; 60: 856–860 13 Shimandle RB, Johnson D, Baker M et al. Safety of Peripheral Intravenous Catheters in Children. Infect Control Hosp Epidemiol 1999; 20: 736–740 14 Vesikari T, Isolauri E, Baer M. A comparative trial of rapid oral and intravenous rehydration in acute diarrhoea. Acta Paediatr Scand 1987; 76: 300–305
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
partially based on studies from low-income countries published in the 1980s or on studies with very small sample sizes. Essentially, the strong recommendation for NGRT mainly refers to a Cochrane review from 2006 (last update in 2010) [7] which cited 17 randomized controlled trials of poor to moderate quality dealing with oral vs. intravenous rehydration in children with gastroenteritis. The route of administration of ORT differed in the cited trials. Only 3 studies have examined sole NGRT vs. IVRT in patients with moderate dehydration secondary to gastroenteritis. The study by Gremse [5] (performed in the US in 1995) included only 24 patients (outpatient setting), the study by Nager [10] (performed in the US in 2002) included 96 patients (outpatient setting) and the study by Sharifi [12] (performed in Iran in 1985) included 470 patients. However, 33 % in the IVRT group and 36 % in the NGRT group of the Sharifi trial were malnourished. In the IVRT group, 26 %, in the NGRT group, 21 % were severely dehydrated [12]. Thus, this study is not representative for high-income countries. There are 6 trials mentioned in the Cochrane review, which analyzed the outcome parameter LOS. Route of oral rehydration solution administration was applied only in one trial by nasogastric tube [5] and in 2 studies by nasogastric tube only when required [3, 9]. Children treated with ORT or NGRT spent significantly less time in hospital ( − 1.20 days) compared to children treated with intravenous fluid. However, this is no longer statistically significant when the outlying study by Gonzalez et al. [3] is removed from the analysis. In this trial, performed in Mexico in 1988, children rehydrated by intravenous fluid spent 4.76 days more in the hospital compared to children who were treated with oral rehydration solution. Since 78 % of the patients in the IVRT group were only mildly dehydrated the question arises why they have a mean LOS of 6.86 days. The LOS is markedly longer than in our study (2.2 ± 1.1 days in the IVRT group) and all other studies in the Cochrane review. 2 other trials showing that the use of ORT decreases LOS are very small in patient numbers (Gremse et al. [5]: 24 patients, − 1.0 days; Vesikari et al. [14]: 37 patients, − 1.2 days), therefore conclusions which can be drawn from these studies are limited. However, these trials, especially the trial by Gonzalez et al., represent the basis for the recommendations for ORT/NGRT in children with acute gastroenteritis concerning the outcome parameter LOS. The largest study performed in a high-income country in which children were rehydrated in the ORT group mainly by mouth and only by NGRT when required showed that LOS was 2 ± 0.71 days in the IVRT group and to 2 ± 1 days in the ORT group [9]. This data is in line with our results suggesting that NGRT compared to IVRT does not reduce LOS in hospitalized children with gastroenteritis. According to the European guidelines [6], ORT and NGRT have significantly fewer adverse events, particularly, the prevention of peripheral intravenous catheters (PVC). It is interesting to note that phlebitis is reported as a common side effect only by Gonzalez et al. (Mexico 1988) [3] and by Sharifi et al. (Iran, 1985) [12]. Our data showed that there were no significant adverse events following IVRT. This data is also in line with different studies in high-income countries showing that peripheral intravenous catheters are safe in children [13]. The adverse event ‘phlebitis’ is mainly reported in publications from low-income countries in the 1980s, suggesting not being representative for high-income countries nowadays.
Nasogastric vs. Intravenous Rehydration in Children with Gastroenteritis and Refusal to Drink: A Randomized Controlled Trial
Authors
J. Marquard1, C. Lerch2, A. Rosen1, H. Wieczorek1, E. Mayatepek1, T. Meißner1
Affiliations
1
Department of General Paediatrics, Neonatology and Paediatric Cardiology, University Children’s Hospital Duesseldorf, Germany 2 Department of General Practice, Cochrane Metabolic and Endocrine Disorders Group, Medical Faculty, University Hospital Duesseldorf, Germany
Key words ▶ gastroenteritis ● ▶ moderate dehydration ● ▶ nasogastric rehydration ● therapy ▶ intravenous rehydration ● therapy ▶ children ●
Abstract
Zusammenfassung
Background: Nasogastric rehydration therapy (NGRT) is the recommended therapy in moderately dehydrated children with gastroenteritis and refusal to drink, since it is supposed to be as effective if not better than intravenous rehydration therapy (IVRT). However, in clinical practice IVRT is often favored. We conducted a clinical trial to determine whether IVRT is not inferior to NGRT. Patients and Methods: Children 3 months to 6 years of age with moderate dehydration and refusal to drink secondary to gastroenteritis were recruited. After clinical assessment of the degree of dehydration, patients were assigned randomly to receive either IVRT or NGRT over 6 h on the hospital ward. Results: Recruitment did not yield the estimated number of patients. Mainly, non-enrollment was due to failure to obtain parental consent because IVRT was expected. 97 patients were enrolled in the study, 46 were randomized to NGRT and 51 to IVRT. There was no difference between IVRT and NGRT groups concerning length of hospital stay (2.2 ± 1.1 days vs. 2.4 ± 1.1 days), success of rehydration (78 vs. 76 %) and adverse events. Discussion: Since we had to terminate the study ahead of schedule due to a low recruiting rate, our results are not reliable. However, data from the literature shows that the widespread described superiority of NGRT over IVRT is seriously influenced by studies from developing countries questioning the applicability of the results to a setting available in high-income countries nowadays. Conclusion: Our study demonstrates the difficulties performing such a study in a high-income country to come to an objective and clearly evident final conclusion.
Hintergrund: Bei Kindern mit moderater Dehydratation, akuter Gastroenteritis (aGE) und Nahrungsverweigerung wird eine nasogastrale Rehydratationstherapie (NGRT) empfohlen. In der klinischen Praxis wird jedoch häufig eine intravenöse Rehydratationstherapie favorisiert (IVRT). Wir untersuchten im Rahmen einer klinischen Studie, ob eine NGRT der IVRT unterlegen ist. Patienten und Methode: Eingeschlossen wurden Patienten im Alter von 3 Monaten bis 6 Jahre mit moderater Dehydratation, aGE und Nahrungsverweigerung. Nach klinischer Einschätzung des Schweregrades der Dehydratation wurden die Patienten randomisiert dem Behandlungsarm NGRT oder IVRT zugeteilt und über 6 h unter stationären Bedingungen rehydriert. Ergebnisse: Die geplante Fallzahl wurde nicht erreicht, viele Eltern lehnten die Studienteilnahme ab, da sie eine IVRT erwarteten. Insgesamt wurde 97 Patienten in die Studie eingeschlossen (46 NGRT, 51 IVRT). Es zeigten sich keine Unterschiede in der Länge des Krankenhausaufenthaltes (2,2 ± 1,1 Tage IVRT vs. 2,4 ± 1,1 Tage NGRT), beim Erfolg der Rehydratationstherapie (78 % IVRT vs. 76 % NGRT) sowie unerwünschten Ereignissen. Diskussion: Da die Studie vorzeitig beendet wurde, können keine Schlussfolgerungen aus den Ergebnissen gezogen werden. Dennoch ist zu erwähnen, dass die häufig beschriebenen Vorteile der NGRT über die IVRT auf Studien aus Entwicklungsländern beruhen und sich nicht auf die heutige Therapie in Industriestaaten übertragen lassen. Schlussfolgerung: Unsere Studie zeigt die Schwierigkeiten der Durchführbarkeit einer Studie in diesem Bereich, um zu klaren Ergebnisse zu kommen.
Schlüsselwörter ▶ Gastroenteritis ● ▶ moderate Dehydratation ● ▶ nasogastrale Rehydratation ● ▶ intravenöse Rehydratation ● ▶ Kinder ●
Bibliography DOI http://dx.doi.org/ 10.1055/s-0033-1363245 Published online: January 16, 2014 Klin Padiatr 2014; 226: 19–23 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0300-8630 Correspondence Dr. Jan Marquard Department of General Paediatrics Neonatology and Paediatric Cardiology University Children’s Hospital Duesseldorf Moorenstraße 5 40225 Düsseldorf Germany Tel.: + 49/211/8117 711 Fax: + 49/211/8119 512 [email protected]
▼
▼
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
Nasogastrale vs. intravenöse Rehydratation bei Kindern mit Gastroenteritis und Nahrungsverweigerung: Eine randomisierte kontrollierte Studie
Background
▼
According to the guidelines of the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition [6] and the American Academy of Pediatrics [8] oral rehydration therapy (ORT) should be used as first-line therapy for the management of gastroenteritis in children. When ORT by mouth fails, nasogastric rehydration therapy (NGRT) should be performed, since it is supposed to be as effective if not better than intravenous rehydration therapy (IVRT) [6]. However, in this situation IVRT is often favored in clinical practice: 75 % of American emergency medicine physicians who indicated familiarity with the guidelines gave IVRT to patients with moderate dehydration [11]. Recently, data from a Canadian pediatric emergency department have been published showing that if oral rehydration fails, only 10 % of caregivers and 14 % of clinicians would choose NGRT [2]. Moreover, caregivers were more familiar with the term intravenous than nasogastric rehydration [2]. This might also be the case for institutions in Europe. Another point is that blood specimen is usually drawn in dehydrated children. Blood withdrawal can be performed with the same procedure the intravenous line is placed. However, when NGRT is the therapy of choice, additional nasogastric tube insertion beside blood withdrawal is always more annoying for the patient. Interestingly, the guidelines for rehydration therapy in children are partially based on studies from low-income countries published in the 1980s or on studies with very small sample sizes. Since we doubt that the studies cited in the guidelines are generally representative for the treatment of moderately dehydrated children with acute gastroenteritis and refusal to drink in high-income countries nowadays, we designed a randomized controlled trial to evaluate the hypothesis that IVRT would not be worse than NGRT. The primary objective of our study was to determine the length of hospital stay (LOS) of moderately dehydrated children rehydrated by nasogastric tube vs. peripheral intravenous catheter (PVC); secondary objectives were to determine the success of rehydration therapy and the occurrence of adverse events.
Methods
▼
Patients and trial design Patients were recruited between August 2008 and December 2011, in the emergency department (ED) of a Children’s University Hospital in Germany. The local Ethics Committee of the Medical Faculty approved the study (approval number 3075). Written informed consent was obtained from all parents. During the study period, all children with gastroenteritis who presented in the emergency department were screened for eligibility. Inclusion criteria were: 1. Age 3 months to 6 years; 2. Acute diarrhea defined as ≥ 3 loose stools of sudden onset in previous 24 h for ≤ 7 days, with or without fever or vomiting [8]; 3. Diagnosis of moderate dehydration using the dehydration score system described by Gorelick et al. (dehydration score ≥ 3 and < 7, corresponding to 5–10 % dehydration) [4] and 4. Refusal to drink, defined as inability of fluid intake tested in the emergency department. The dehydration score by Gorelick includes the parameter “Decreased skin elasticity”, “Capillary refill > 2 s”, “General appearance”, “Absent tears”, “Abnormal respiration”, “Dry mucous membranes”, “Sunken eyes”, “Abnormal radial pulse”, “Tachycardia (heart rate > 150 beats per min)” and “Decreased urine output” [4].
Exclusion criteria were significant other morbidity, pre-treatment in any emergency department or hospital in the preceding 24 h and any injury or abnormality in the nasopharyngeal space. Patients were randomized, as described in detail below, to receive either IVRT or NGRT. For children randomized to IVRT, an intravenous line was placed by a physician, in the NGRT group a nasogastric tube was inserted by a nurse. To make sure that the nasogastric tube is at the right place fluid drawn out from the tube was tested for acidity with a pH indicator to show whether or not the tube is in the stomach. A maximum of 3 attempts at intravenous line or nasogastric tube insertion was allowed. If more than 3 attempts failed, physicians respectively nurses were recommended to cross over to the other treatment. Blood was collected in both groups (intravenous blood collection in the IVRT, capillary blood collection in the NGRT group) to perform a blood gas analysis including blood electrolytes and blood glucose. For rehydration therapy, the patients were moved to the ward. A pre-therapy naked weight in children with a body weight less than 10 kg was measured to the nearest 0.01 kg using a SECA electronic baby scale by nursing staff, in children with a body weight more than 10 kg body weight in light indoor clothes was measured to the nearest 0.1 kg using a SECA chair scale [SECA, Hamburg, Germany]. Patients were rehydrated during a 6-h rehydration phase: in the IVRT group they received 50 ml/kg of a 2.5 % glucose in 0.45 % sodium chloride solution over 6 h via an infusion pump. In the NGRT group they received the same amount of oral rehydration solution containing 60 mmol/L sodium, 20 mmol/L potassium, 60 mmol/L chloride, 10 mmol/L citrate and 90 mmol/L glucose over 6 h at a constant rate through a nasogastric tube. In our hospital, the use of oral ondansetron is uncommon in children with gastroenteritis and was not used in the study. The local Ethics Committee recommended the use of racecadotril in all study patients because it was a standard treatment for children with gastroenteritis with 3 or more watery or loose stools within 24-h in our hospital. The dose was 1.5 mg/kg, 3 times daily at regular intervals until 2 normal stools were recorded. In the NGRT group racecadotril was administered via the nasogastric tube, in the IVRT group children had to drink racecadotril, added to water. Since acute diarrhea defined as ≥ 3 loose stools of sudden onset in previous 24 h was an inclusion criterion, all patients received racecadotril. However, since refusal to drink was an inclusion criterion as well, the intake of racecadotril was more common in the NGRT group. This has seriously influenced the study in favor for NGRT. After 6 h of rehydration, patients were reassessed according to the dehydration score system [8] for signs of dehydration, and body weight was measured again, under the same conditions as prior to treatment. If refusal to drink was ongoing and or rehydration was considered as “not successful”, maintenance fluid requirement was administered according to the formula: First 10 kg 100 ml/kg/day, second 10 kg 50 ml/kg/day and subsequent kg 20 ml/kg/day. If necessary, physicians adjusted the rate of rehydration to replace the volume of ongoing losses. During the hospitalization, oral fluid intake, number of vomiting, and diarrhea were measured constantly. Patients were discharged once oral feeding was tolerated (defined as oral fluid intake > 25 % of daily recommended fluid amount in last 6 h), in the absence of clinical signs of dehydration and vomiting, and less than 5 loose stools in the last 24 h. A follow-up visit was not performed within this study.
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
20 Original Article
Original Article 21
Fig. 1 Flow diagram. Assessed for eligibility (n= 4 715)
Excluded (n= 4618) ◆ Not meeting inclusion criteria (n= 4439) ◆ Declined to participate (n= 179)
Allocated to NGRT (n= 46) ◆ Received allocated intervention (n= 43) ◆ Did not receive allocated intervention (n= 3) (unable to obtain nasogastric intubation)
Allocated to IVRT (n= 51) ◆ Received allocated intervention (n= 51) ◆ Did not receive allocated intervention (n= 0)
◆ Lost to follow-up (n= 0) ◆ Discontinued intervention (n= 12) (persistent vomiting)
◆ Lost to follow-up (n= 0) ◆ Discontinued intervention (n= 0)
46 analyzed (intention-to-treat)
51 analyzed (intention-to-treat)
Outcome measurements The primary outcome was LOS in days, calculated by subtracting day of admission from day of discharge (equivalent to number of nights). Minimum LOS was 1 day (1 night). LOS in hours was not calculated since discharge of patients was performed within the normal routine during and day and therefore could have biased the results. Secondary outcomes included success of rehydration therapy, defined as resolution of moderate dehydration (a 6-h dehydration score ≤ 2, weight gain and production of urine (defined as urine in nappy, urine volume was not determined) within the rehydration phase). NGRT was changed to IVRT when treatment was not tolerated, e. g. due to persistent vomiting (defined as ≥ 3 episodes of vomiting). Occurrence of adverse events was defined as diagnosis of therapy related events like phlebitis or extravasation injuries in the IVRT group. Body temperature, vital signs, urine volume and number of placement failures in the IVRT or NGRT group were not analyzed in this study. Parameters of pain and dissatisfaction of the patients were not documented. Cost effectiveness of both procedures was not considered as outcome measure.
sequence that was stratified by age ( ≤ 3 years vs. > 3 years) and applied 3 different random block sizes. Sequentially numbered, opaque and sealed envelopes concealed the allocation sequence from the clinicians.
Data analyses Data was recorded on a case record form. All analyses followed the intention to treat principle and included patients with protocol deviations. Analyses were performed with SPSS software (version 11.5.1, 2002). Data was presented as means ± standard deviation (SD), medians with interquartile range (IQR), absolute numbers or specified in percent. Categorical data was analyzed by Chi-Square test, continuous data was analyzed by MannWhitney test. For the primary outcome parameter LOS we also calculated the difference in means and its 95 % confidence interval using a 2-sample t-test. A p-value of < 0.05 was considered as statistically significant.
Results
▼
Patients Sample-size calculation Data from our hospital indicated that the standard deviation of LOS was 1.5 days. To detect a reduction in LOS of 0.6 days (that is, 3 out of 5 children would be spared one night in hospital) with a 2-sided α of 5 %, a power of 80 % and an anticipated dropout rate of 10 %, a sample size of 110 patients per group was necessary [1].
Randomization Allocation of the participants occurred according to a computer generated randomization list. Stata 10 (College Station, Texas, USA: StataCorp LP) was used for creating the 1:1 randomization
During the study period, 4 715 patients presented to the emergency department with clinical signs of acute gastroenteritis. 276 eligible patients were identified of whom 97 (35.1 %) were enrolled in the study, 46 patients were allocated to NGT and 51 to IVRT. Caregivers of 179 patients (64.9 %) declined to participate. In 3 patients of the NGT group, placement of nasogastric tube ▶ Fig. 1). Baseline characfailed (6.5 %) and IVRT was performed (● ▶ Table 1). teristics were similar between the study groups (● The recruitment did not yield the number of patients estimated in the sample size calculation. Since we had the impression that there was an increased rate of change from NGRT to IVRT due to vomiting, we decided to conduct a non pre-specified interim
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
Randomized (n= 97)
22 Original Article
analysis concerning ‘Allocated intervention not tolerated’ after 97 patients were included. The analysis showed that change from NGRT to IVRT was significant and we decided to end ahead of schedule. Another critical point was a recruitment rate lower than anticipated. Reason for non-enrollment of eligible patients was failure to obtain parental consent. Only the intention to treat analysis is reported since the results were similar to those of the per-protocol analysis.
ing rehydration (clinical diagnosis made by the ward physician) and 8 did not tolerate the therapy due to persistent vomiting (17 %). The overall rate of change of rehydration therapy was significantly more common in the NGRT group compared to the IVRT group (26 vs. 0 %, p < 0.001). One patient in the IVRT group had a mild extravasation injury, there was no patients with phle▶ Table 2). In the NGRT group no adverse events were bitis (● reported during the treatment phase (e. g. misplacement of tube, removal by the child).
Primary outcome
Secondary outcomes ▶ Table 2 gives details about secondary outcomes. No difference ● between the 2 groups in overall success of rehydration at 6 h could be found (78 % in the IVRT group and 76 % in the NGRT group). Resolution of moderate dehydration and weight gain was found to be the same in both groups. However, production of urine was statistically more common in the IVRT group compared to the NGRT group (98 % vv. 90 %, p = 0.035). Of the 46 patients in the NGRT group, 4 (9 %) had clinical impairment durTable 1 Baseline characteristics of the study sample. Variable age (years) mean median sex ( % female) baseline dehydration score mean median blood base excess (mmol/l) median blood ph median blood sodium (mmol/l) median blood potassium (mmol/l) median leucocytes (x109/l) median
IVRT (n = 51)
NGRT (n = 46)
1.7 ± 1.1 1.3 (0.8–2.3) 41 (n = 21)
1.8 ± 1.1 1.4 (1.0–2.3) 43 (n = 20)
3.6 ± 0.7 3.0 (3.0–4.0)
3.6 ± 0.7 3.5 (3.0–4.0)
− 7.3 ( − 9.8 to − 5.1)
− 8.1 ( − 9.3 to − 5.1)
7.37 (7.36–7.40)
7.38 (7.36–7.40)
136 (135–138)
138 (136–139)
3.9 (3.6–4.3)
4.2 (3.8–4.5)
8.6 (5.9–11.9)
7.8 (7.1–10)
Data are shown for the intention-to-treat population. Data are given as median with interquartile range or specified in percent. IVRT intravenous rehydration therapy, NGRT nasogastric rehydration therapy
Variable length of hospital stay (days) mean median successful rehydration ( %) resolution of moderate dehydration (score ≤ 2) ( %) production of urine ( %) weight gain ( %) change of rehydration therapy ( %) clinical impairment during rehydration ( %) allocated intervention not tolerated ( %) complications phlebitis mild extravasation injury ( %)
Discussion
▼
After termination the study ahead of schedule, results of the interim analysis show no difference in LOS, success of rehydration therapy, and occurrence of adverse events between IVRT and NGRT in moderately dehydrated children with acute gastroenteritis and refusal to drink in the setting of our hospital. However, it has to be mentioned that we cannot draw any conclusions out of these data since the recruitment did not yield the number of patients estimated in the sample size calculation. Interestingly, the recruitment rate was dramatically lower than anticipated. Caregivers of eligible patients were informed on both rehydration regimes. If they declined to participate, they could still choose between the 2 options (NGRT or IVRT). Remarkably, all non-enrolled eligible patients were rehydrated intravenously, suggesting parental preference for IVRT. Reason for non-enrollment of eligible patients was failure to obtain parental consent because caregivers expected IVRT, were familiar with IVRT or believed that NGRT is more painful. However, IVRT was the established rehydration therapy in children with gastroenteritis in our hospital prior to this study, so the physicians’ tendency to IVRT could also have influenced parents in their decision on therapy regime or on willingness to participate in this study. This finding is similar to Freedmans et al. [2] results showing that caregivers are more familiar with the term IVRT and believe that intravenous insertion is easier compared to nasogastric insertion. Despite termination the study ahead of schedule, there is one important finding: Parental and physician’s preference for IVRT seems to be tremendously high in Germany and probably in other high-income countries as well. Therefore the feasibility of our study design is called into question. However, we still see a rationale for the question whether IVRT should also be a recommended therapy in high income countries since the guidelines for rehydration therapy in children are
IVRT (n = 51)
NGRT (n = 46)
p-Value
2.2 ± 1.2 2 (1–3) 78 (40/51) 92 (47/51) 98 (50/51) 96 (49/51) 0 (0/51) 0 (0/51) 0 (0/51)
2.4 ± 1.2 2 (1.3–3) 76 (35/46) 90 (40/46) 90 (40/46) 91 (42/46) 26 (12/46) 9 (4/46) 17 (8/46)
0.34 0.78 0.4 0.035 0.33 < 0.001 0.031 0.002
0 (0/51) 2 (1/51)
0 (0/46) 0 (0/46)
Data are given as means ± SD, median with interquartile range or specified in percent
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
0.34
Table 2 Primary and secondary outcome measurements.
Downloaded by: National University of Singapore. Copyrighted material.
There was no difference between the 2 groups in terms of LOS: 2.2 ± 1.1 days in the IVRT group and 2.4 ± 1.1 days in the NGRT ▶ Table 2) (difference in LOS − 0.2 days; 95 % confidence group (● interval, − 0.689–0.3036; p = 0.44).
Original Article 23 Conclusion
▼
There is no doubt that ORT by mouth is the treatment of choice in dehydrated children with gastroenteritis. If children are able to drink oral rehydration solution they do not require a nasogastric tube or peripheral venous catheter, both potentially painful and difficult procedures. When ORT by mouth fails, NGRT is the recommended therapy in Germany. Although rehydration is a frequent daily procedure in German pediatric hospitals, there are no reliable data for Germany or high-income countries available showing that NGRT is superior to IVRT for rehydration of moderately dehydrated children with gastroenteritis and refusal to drink. Since we had to terminate the study ahead of schedule due to a low recruiting rate, we are not able to answer this question reliable based our results. However, parental and physician’s preference for IVRT seems to be tremendously high in Germany and probably in other high-income countries as well. Our study demonstrates the difficulties performing such a study in a high-income country to come to an objective and clearly evident final conclusion.
Conflict of interest: The authors have no conflict of interest to disclose. References 1 Dupont WD, Plummer WD. Power and Sample Size Calculations: A Review and Computer Program. Control Clin Trials 1990; 11: 116–128 2 Freedman SB, Keating LE, Rumatir M et al. Health Care Provider and Caregiver Preferences Regarding Nasogastric and Intravenous Rehydration. Pediatrics 2012; 130: e1504–e1511 3 Gonzalez-Adriano SR, Valdes-Garza HE, Garcia-Valdes LC. Oral hydration versus intravenous hydration in patients with acute diarrhea. Bol Med Hosp Infant Mex 1988; 45: 165–172 4 Gorelick MH, Shaw KN, Murphy KO. Validity and reliability of clinical signs in the diagnosis of dehydration in children. Pediatrics 1997; 99: E6 5 Gremse DA. Effectiveness of nasogastric rehydration in hospatalized children with acute diarrhea. JPGN 1995; 21: 145–148 6 Guarino A, Albano F, Ashkenazi S et al. European Society for Paediatric Gastroenterology, Hepatology, and Nutrition/European Society for Paediatric Infectious Diseases: Evidence-based Guidelines for the Management of Acute Gastroenteritis in Children. JPGN 2008; 46: S81–S122 7 Hartling L, Bellemare S, Wiebe N et al. Oral versus intravenous rehydration for treating dehydration due to gastroenteritis in children. Cochrane Database Syst Rev 2006; 3: CD004390 8 King CK, Glass R, Bresee JS et al. Centers for Disease Control and Prevention: Managing acute gastroenteritis among children: oral rehydration, maintenance, and nutritional therapy. MMWR Recomm Rep 2003; 21: 1–16 9 Mackenzie A, Barnes G. Randomised controlled trial comparing oral and intravenous rehydration therapy in children with diarrhoea. BMJ 1991; 303: 393–396 10 Nager AL, Wang VJ. Comparison of nasogastric and intravenous methods of rehydration in pediatric patients with acute dehydration. Pediatrics 2002; 109: 566–572 11 Ozuah PO, Avner JR, Stein RE. Oral rehydration, emergency physicians, and practice parameters: a national survey. Pediatrics 2002; 109: 259–261 12 Sharifi J, Ghavami F, Nowrouzi Z et al. Oral versus intravenous rehydration therapy in severe gastroenteritis. Arch Dis Child 1985; 60: 856–860 13 Shimandle RB, Johnson D, Baker M et al. Safety of Peripheral Intravenous Catheters in Children. Infect Control Hosp Epidemiol 1999; 20: 736–740 14 Vesikari T, Isolauri E, Baer M. A comparative trial of rapid oral and intravenous rehydration in acute diarrhoea. Acta Paediatr Scand 1987; 76: 300–305
Marquard J et al. Nasogastric vs. Intravenous Rehydration … Klin Padiatr 2014; 226: 19–23
Downloaded by: National University of Singapore. Copyrighted material.
partially based on studies from low-income countries published in the 1980s or on studies with very small sample sizes. Essentially, the strong recommendation for NGRT mainly refers to a Cochrane review from 2006 (last update in 2010) [7] which cited 17 randomized controlled trials of poor to moderate quality dealing with oral vs. intravenous rehydration in children with gastroenteritis. The route of administration of ORT differed in the cited trials. Only 3 studies have examined sole NGRT vs. IVRT in patients with moderate dehydration secondary to gastroenteritis. The study by Gremse [5] (performed in the US in 1995) included only 24 patients (outpatient setting), the study by Nager [10] (performed in the US in 2002) included 96 patients (outpatient setting) and the study by Sharifi [12] (performed in Iran in 1985) included 470 patients. However, 33 % in the IVRT group and 36 % in the NGRT group of the Sharifi trial were malnourished. In the IVRT group, 26 %, in the NGRT group, 21 % were severely dehydrated [12]. Thus, this study is not representative for high-income countries. There are 6 trials mentioned in the Cochrane review, which analyzed the outcome parameter LOS. Route of oral rehydration solution administration was applied only in one trial by nasogastric tube [5] and in 2 studies by nasogastric tube only when required [3, 9]. Children treated with ORT or NGRT spent significantly less time in hospital ( − 1.20 days) compared to children treated with intravenous fluid. However, this is no longer statistically significant when the outlying study by Gonzalez et al. [3] is removed from the analysis. In this trial, performed in Mexico in 1988, children rehydrated by intravenous fluid spent 4.76 days more in the hospital compared to children who were treated with oral rehydration solution. Since 78 % of the patients in the IVRT group were only mildly dehydrated the question arises why they have a mean LOS of 6.86 days. The LOS is markedly longer than in our study (2.2 ± 1.1 days in the IVRT group) and all other studies in the Cochrane review. 2 other trials showing that the use of ORT decreases LOS are very small in patient numbers (Gremse et al. [5]: 24 patients, − 1.0 days; Vesikari et al. [14]: 37 patients, − 1.2 days), therefore conclusions which can be drawn from these studies are limited. However, these trials, especially the trial by Gonzalez et al., represent the basis for the recommendations for ORT/NGRT in children with acute gastroenteritis concerning the outcome parameter LOS. The largest study performed in a high-income country in which children were rehydrated in the ORT group mainly by mouth and only by NGRT when required showed that LOS was 2 ± 0.71 days in the IVRT group and to 2 ± 1 days in the ORT group [9]. This data is in line with our results suggesting that NGRT compared to IVRT does not reduce LOS in hospitalized children with gastroenteritis. According to the European guidelines [6], ORT and NGRT have significantly fewer adverse events, particularly, the prevention of peripheral intravenous catheters (PVC). It is interesting to note that phlebitis is reported as a common side effect only by Gonzalez et al. (Mexico 1988) [3] and by Sharifi et al. (Iran, 1985) [12]. Our data showed that there were no significant adverse events following IVRT. This data is also in line with different studies in high-income countries showing that peripheral intravenous catheters are safe in children [13]. The adverse event ‘phlebitis’ is mainly reported in publications from low-income countries in the 1980s, suggesting not being representative for high-income countries nowadays.
