Alternative Rehydration Methods: A Systematic Review and Lessons for Resource-Limited Care Shada Rouhani, Laura Meloney, Roy

Ahn, Brett D. Nelson and Thomas F. Burke Pediatrics 2011;127;e748; originally published online February 14, 2011; DOI: 10.1542/peds.2010-0952

The online version of this article, along with updated information and services, is located on the World Wide Web at:
http://pediatrics.aappublications.org/content/127/3/e748.full.html

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007. Copyright 2011 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org by guest on October 17, 2013

Alternative Rehydration Methods: A Systematic Review and Lessons for Resource-Limited Care Shada Rouhani, Laura Meloney, Roy Ahn, Brett D. Nelson and Thomas F. Burke Pediatrics 2011;127;e748; originally published online February 14, 2011; DOI: 10.1542/peds.2010-0952
Updated Information & Services References including high resolution figures, can be found at: http://pediatrics.aappublications.org/content/127/3/e748.full.h tml This article cites 46 articles, 11 of which can be accessed free at: http://pediatrics.aappublications.org/content/127/3/e748.full.h tml#ref-list-1 This article has been cited by 2 HighWire-hosted articles: http://pediatrics.aappublications.org/content/127/3/e748.full.h tml#related-urls This article, along with others on similar topics, appears in the following collection(s): Complementary & Integrative Medicine http://pediatrics.aappublications.org/cgi/collection/compleme ntary_-_integrative_medicine_sub Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://pediatrics.aappublications.org/site/misc/Permissions.xht ml Information about ordering reprints can be found online: http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Citations

Subspecialty Collections

Permissions & Licensing

Reprints

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007. Copyright 2011 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from pediatrics.aappublications.org by guest on October 17, 2013

Alternative Rehydration Methods: A Systematic Review and Lessons for Resource-Limited Care abstract
OBJECTIVE: Dehydration is a signicant threat to the health of children worldwide and a major cause of death in resource-scarce settings. Although multiple studies have revealed that oral and intravenous (IV) methods for rehydration in nonsevere dehydration are nearly equally effective, little is known about effectiveness beyond these 2 techniques. With this systematic review we analyzed the effectiveness of nonoral and nonintravenous methods of rehydration. METHODS: The Medline, Cochrane, Global Health, Embase, and CINAHL (Cumulative Index to Nursing and Allied Health Literature) databases were searched for articles on intraosseous (IO), nasogastric (NG), intraperitoneal (IP), subcutaneous (hypodermoclysis), and rectal (proctoclysis) rehydration through December 2009. Only human pediatric studies that included data on the effectiveness or complications of these methods were included. RESULTS: The search identied 38 articles that met the inclusion criteria: 12 articles on NG, 16 on IO, 7 on IP, 3 on subcutaneous, and none on rectal rehydration. NG rehydration was as effective as IV rehydration for moderate-to-severe dehydration. IO rehydration was effective and easy to obtain, although only 1 randomized trial was identied. IP rehydration had some benet for moderate dehydration, although none of the trials had control groups. Limited data were available on subcutaneous rehydration, and only 1 case series showed benet. CONCLUSIONS: NG rehydration should be considered second-line therapy, after oral rehydration, particularly in resource-limited environments. IO rehydration seems to be an effective alternative when IV access is not readily obtainable. Additional evidence is needed before IP and subcutaneous rehydration can be endorsed. Pediatrics 2011; 127:e748e757
AUTHORS: Shada Rouhani, MD,a Laura Meloney, BS,b Roy Ahn, SD,b,c Brett D. Nelson, MD, MPH, DTM&H,b,d,e and Thomas F. Burke, MDb,c,d,e
aHarvard Afliated Emergency Medicine Residency Program, Brigham and Womens Hospital and Massachusetts General Hospital, Boston, MA; bDivision of Global Health and Human Rights, Department of Emergency Medicine, Massachusetts General Hospital; Boston, MA; Departments of cSurgery and ePediatrics, Harvard Medical School, Boston, MA; dDivision of Global Health; MassGeneral Hospital for Children, Boston, MA

KEY WORDS dehydration, critically ill children, diarrhea, international child health, intravenous rehydration ABBREVIATIONS IVintravenous NGnasogastric IOintraosseous IPintraperitoneal ORSoral rehydration solution NSnormal saline www.pediatrics.org/cgi/doi/10.1542/peds.2010-0952 doi:10.1542/peds.2010-0952 Accepted for publication Nov 19, 2010 Address correspondence to Shada Rouhani, MD, Department of Emergency Medicine, Massachusetts General Hospital, 5 Emerson Place, Suite 100, Boston, MA 02114. E-mail: srouhani@ partners.org PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright 2011 by the American Academy of Pediatrics FINANCIAL DISCLOSURE: The authors have indicated they have no nancial relationships relevant to this article to disclose.

e748

ROUHANI et al

REVIEW ARTICLES

Dehydration is one of the most common problems confronting ill children worldwide. It is closely tied to many of the leading causes of childhood mortality, including diarrheal illnesses, acute respiratory infections, malaria, and malnutrition. The United Nations Childrens Fund has estimated that 1.7 million children per year die from diarrheal dehydration alone.1 It is well established that when available and tolerated, oral rehydration is the preferred method of rehydration. It is as effective as intravenous (IV) rehydration in mild and moderate dehydration and is more readily available and affordable.2,3 However, for some patients oral rehydration is not possible because of unwillingness or inability to drink, severity of illness, or comorbidities. In resource-scarce settings, sterile supplies needed to secure and maintain IV access are often absent. Even if supplies are available, children in the developing world frequently present with such severe dehydration that intravascular access is technically challenging or impossible. Because oral and IV rehydration are not feasible for all children, it is important for clinicians to have alternative methods of rehydration available to them. Numerous alternative methods have been used. Nasogastric (NG) rehydration uses the intestinal system just as oral rehydration does. Proctoclysis relies on the absorption of uids through the rectal mucosa. Intraosseous (IO) uid administration uses bone marrow as a direct access to the intravascular circulation. Hypodermoclysis, or subcutaneous uid administration, and intraperitoneal (IP) rehydration both rely on indirect absorption into the circulation. The relative effectiveness of these methods for acute rehydration is unclear. Some, such as NG and IO rehydration, are used today at the discretion of the
PEDIATRICS Volume 127, Number 3, March 2011

practitioner. Others are not widely used. The aim of this review was to examine the literature on the effectiveness, benets, and risks of alternative (ie, nonoral and non-IV) techniques for pediatric rehydration. Techniques examined included NG, IO, IP, subcutaneous, and proctoclysis rehydration.

METHODS
A literature search was conducted in 5 databases: Medline, Cochrane, Global Health, Embase, and CINAHL (Cumulative Index to Nursing and Allied Health Literature). The following text search terms and Medical Subject Headings were used: rehydration; dehydration; hydration; uid therapy; and injection, in combination with any of NG, IO, IP, subcutaneous, hypodermoclysis, or proctoclysis. The searches were limited to articles written in English and concerning human studies. The searches were not restricted by dates of publication and included all articles through the completion of the search in December 2009. In addition to the articles identied through this search, key references from included articles were also examined. The identied articles were manually sorted to include patients younger than 18 years. If studies had both pediatric and adult populations, they were included only if they had listed the results separately for the pediatric patients. Adult studies were excluded, because much of the adult data on alternative rehydration methods, such as subcutaneous rehydration, come from geriatric and palliative care literature. These patients have different physiology and goals of care from children with acute dehydration. Case reports and consensus-based guidelines were eligible for inclusion, whereas articles that were opinion-based were excluded. Articles in which a technique but not its effectiveness was described

were also excluded. However, articles in which a complication of a method was described, regardless of reference to effectiveness, were included to allow an assessment of risks and benets. The search strategy is outlined in Fig 1. The selection of articles was initially conducted by 1 author; when an article was felt to be borderline for inclusion or exclusion, it was examined by 2 or more authors, and a consensus on inclusion or exclusion was reached. Given the limited literature on this subject, articles that met the abovelisted inclusion criteria were not excluded on the basis of the quality of their methods, but assessment of the differences in methodology was incorporated into the text of the review to allow the reader to weigh the quality of the literature.

RESULTS
The initial search strategy identied 1436 articles. A signicant number of these articles were excluded, because they focused solely on adult populations. In addition, many studies did not relate to dehydration or hypovolemia. After manual review of the titles and available abstracts of these articles, the list was narrowed to 82 articles. Subsequent review of the full text of these articles identied 38 articles that met all inclusion criteria: 3 on subcutaneous rehydration, 12 on NG rehydration, 7 on IP rehydration, 16 on IO rehydration, and none on proctoclysis. NG Rehydration Twelve articles about NG hydration met the inclusion criteria for our review. Of these studies, 5 were randomized controlled trials, 4 of which compared NG to IV rehydration and 1 of which compared different solutions for NG rehydration. The remaining articles included cohort studies, case series, retrospective chart reviews, and consensus-panel statements. The articles reviewed suggested that NG rehye749

Database search:
Medline, Cochrane, Global Health, Embase, CINAHL
Limits: Human studies English language No date restricon

Manual sorng
Exclude: Adult-only results Opinion-based reports Menons technique but not eecveness

be switched to IV rehydration.10,11 Two other case series revealed benets with NG rehydration consistent with those in other studies, although neither the level of dehydration nor the rehydration protocol that was followed were specied.12,13 Finally, one consensus guideline on the management of diarrhea with or without vomiting addressed NG rehydration. It was developed by using a Delphi consensus process after a systematic literature review. The panel recommended NG rehydration over IV rehydration for patients in whom oral rehydration had failed.14 Complications from NG rehydration were infrequent. One study found that emesis was more common in patients with gastroenteritis who were given NG rehydration compared with IV rehydration.15 One patient developed meteorism, or excessive gas accumulation in the gastrointestinal tract, and was successfully switched to IV rehydration.5 In a study on adverse events in patients with NG tubes for rehydration, reported adverse effects included multiple insertions in 34%, sore throat in 13%, coughing in 16%, negligible epistaxis in 3%, and need for mittens in nearly 75% of the patients.16 However, in this study, data on adverse events were collected from only 26% of the eligible patients. There was only 1 aspiration event in any of the included studies, and after review of the chart in that case, the authors concluded that it was unclear if the recorded event was aspiration or misplacement. IO Rehydration Sixteen articles addressed IO infusions. One was a randomized controlled trial, 12 were case reports, and 3 were case series. Of the 12 case reports, 10 described the efcacy of IO infusion and 2 described complications. The randomized controlled trial compared IV to IO hydration in severely de-

Include: All arcles with children <18 y old Menon of eecveness or complicaons

38 arcles 3 hypodermoclysis subcutaneous 12 NG 16 IO

0 proctoclysis

7 IP

FIGURE 1
Schematic diagram of the literature search and inclusion/exclusion criteria.

dration is effective in cases of moderate and severe dehydration. The 4 randomized trials that compared NG to IV rehydration are summarized in Table 1. In all 4 studies, NG rehydration had efcacy similar to that of IV rehydration, although different end points and protocols were used in each study. NG rehydration failed in 2% of all patients (4 of 211 total) for reasons detailed in Table 1. These patients were switched to IV rehydration. Several case series also supported the effectiveness of NG rehydration. In 2000, a prospective study of 4131 children with acute diarrhea and severe dehydration, dened as at least a 10% weight loss, supported NG rehydration. All children were rehydrated by NG tube using oral rehydration solution (ORS) at 20 to 30 drops per minute and a total volume of 50 to 100 mL/kg per day for the rst 10 kg of weight and 25 to 50 mL/kg per day for the remaining
e750 ROUHANI et al

weight. Ninety percent of the patients had signicant weight gain after 4 hours of NG infusion, 10% needed prolonged NG infusion, and none required IV uids.8 A subsequent study in 2003 included 4000 children with acute diarrhea and a 10% weight loss, although patients in shock were excluded. Patients were randomly assigned to receive NG rehydration solutions of different osmolarity. Of these children, 3537 (88%) had a rapid response to uids, 316 (8%) needed prolonged rehydration, and 147 (4%) needed admission to the hospital.9 It is unclear if children admitted to the hospital were given IV rehydration. Smaller case series also supported the effectiveness of NG rehydration. In 2 studies focused on children with moderate dehydration, most patients experienced clinical improvement with NG rehydration, although 4 of 47 patients (9%) in the 2 studies needed to

TABLE 1 Randomized Controlled Trials in Which IV and NG Rehydration Were Compared

Infusion Fluids NG: Rehydralyte (ORS), volume to replace uid decit over 6 h; IV: D512NS, volume not specied Weight; intakes and outputs; frequency and volume of vomiting and diarrhea; duration of hospitalization; serum electrolyte levels End Points Results

Author (Year)

Population

PEDIATRICS Volume 127, Number 3, March 2011 NG: WHO standard ORS; IV: lactated Ringers solution, 40 mL/kg 1 h, then 30 mL/kg 1 h, then 20 mL/kg 2 h Weight; hourly assessment of degree of dehydration; 4-h assessment of degree of hydration 50 mL/kg of Pedialyte by NG tube or NS by IV line over 3 h, then switched to oral uids Vital signs; general clinical state; capillary rell; change in weight; change in serum electrolyte levels; urine chemistries; urine specic gravity; telephone follow-up; returnvisit rates 150 mL/kg per d (1020 mL/kg per h 2 h, then a constant rate over 22 h); NG: bedside ORS; IV: D513NS Weight; blood and urine samples; plasma specic gravity Rehydration successful in 11 of 12 patients receiving NG rehydration; failed for 1 patient because of persistent vomiting; Patients in the NG group had formula introduced sooner (at discretion of attending) and had shorter hospitalizations; No complications of NG placement; most frequent IV complication was IV inltration (patients needed an average of 1.9 IV catheters) No differences in rates of successful rehydration, recurrence of dehydration, mean hospital stay, or duration of vomiting; Rehydration succeeded for 33 of 36 children in the NG group (failed for 2 patients because of persistent vomiting and for 1 because of meteorism, or excessive gas accumulation in the gastrointestinal tract) and 37 of 39 children in the IV group (failed for 1 because of persistent vomiting and for 1 because of seizures); differences were not statistically signicant No difference in discharge rates, clinical condition at follow-up, or number of return visits; both groups gained weight, although those in the NG group gained less as a percentage of body weight (2.2% vs 3.6%); IV placement failed more frequently than NG placement (61% vs 4% mean per-case failure rate) Similar weight gain, oral intake, and improvement in laboratory test results between groups; sodium and osmolarity returned to normal without complications in both groups

Gremse4 (1995)

224 mo old; 24 patients; moderate dehydration; developed country

Hidayat et al5 (1988)

159 mo old; 75 patients; severe dehydration; developing country; at study onset, the NG group had a similar duration of illness to the IV group, but fewer episodes of vomiting per day

Nager and Wang6 (2002)

336 mo old; 90 childrena; moderate dehydration; emergency department patients; developed country

Varavithya et al7 (1978)

417 mo old; males; 22 children; moderate dehydration; developing country

All studies were of children with acute dehydration secondary to vomiting or diarrhea. Gremse4 and Nager and Wang6 included children for whom an oral challenge had failed. In all articles, moderate dehydration was estimated to be 5% to 10%. Nager and Wang6 and Varavithya et al7 both dened signs of moderate dehydration as tachycardia, decreased skin turgor, sunken fontanelles and eyes, dry mucous membranes, delayed capillary rell, and normal or listless mental status. Severe dehydration was dened by Hidayat et al5 by using World Health Organization criteria and was estimated to be 10% dehydration. These criteria included lethargy or unconsciousness, very decreased skin turgor, very dry mucous membranes, no tears, and very sunken eyes. Gremse4 did not specify how the degree of dehydration was estimated. Exclusion criteria varied between studies. Hidayat et al5 excluded patients with meteorism, or excessive gas accumulation, complications (not dened), and a nonpalpable pulse; Gremse4 excluded shock, sepsis, ileus, seizures, metabolic disease, intestinal obstruction, and chronic disease; Nager and Wang6 excluded children with severe dehydration, shock, intractable vomiting, underlying diseases, or suspected other acute etiology (such as appendicitis, malrotation, or meningitis); and Varavithya et al7 did not list any exclusion criteria. D5NS indicates 5% dextrose NS; D512NS, 5% dextrose 0.5% NS; D513NS, 5% dextrose 0.3% NS; WHO, World Health Organization. a The Nager and Wang6 study initially enrolled 96 patients; 3 were withdrawn because of continued emesis that required hospital admission (2 IV, 1 NG); 1 was withdrawn because of intussusception, and 2 were withdrawn after they were found to have severe dehydration in retrospect (both in the IV group).

REVIEW ARTICLES

e751

hydrated children with acute gastroenteritis.17 Sixty children received a 20 to 30 mL/kg normal saline (NS) bolus by IO or IV routes, followed by identical protocols regarding the reintroduction of oral uids. End points included time to placement, stabilization of vital signs, correction of dehydration, and complications. There was no difference in efcacy of rehydration or correction of laboratory abnormalities between the groups. However, IO placement was signicantly faster (67 vs 129 seconds) and more reliable (IV lines failed to be placed within 5 minutes in 33% of the patients in the IV group, versus no failures in the IO group). IO needles were successfully placed in those patients in the IV group in whom IV placement had failed; it is unclear in which group these patients were subsequently analyzed. There were no short-term complications in either group, although the study lacked long-term follow-up. Results of another study of 22 children aged 1.5 to 10 months with shock further supported the efcacy of IO uids for rehydration.18 Children had an IO line placed either immediately if they had gasping respirations or after 2 to 3 failed IV-line-placement attempts. Twenty-one of 22 patients (95.5%) had improvement in their circulatory state with uids, blood, and medications given through the IO route. Sixteen (73%) of the patients ultimately survived. The only patient with no response to uids was apneic and bradycardic on arrival. It is important to note that in all patients, IO access was obtained by using a standard 18-gauge buttery needle rather than an IOspecic needle. Four case reports focused on IO-uid resuscitation in patients with burns. A total of 6 patients (age range: 17 days to 3 years) were included in these reports.1921 All 6 children were successfully initially resuscitated with IO
e752 ROUHANI et al

uid, and 2 continued to receive IO uids for 48 hours.20 One child died from smoke inhalation 2 days after the initial resuscitation.21 Two case reports involved uids given through the IO route during cardiac arrest. One child was successfully resuscitated after receiving lactated Ringers solution and multiple vasoactive medications.22 The other child did not survive, but a femoral vein puncture above the site of the IO infusion produced diluted blood that appeared to be mixed with saline, which provided evidence that the IO-infused saline had reached the intravascular space.23 Two additional case reports on hypovolemic patients who were resuscitated with IO access reported successful outcomes. One was a 5-year-old patient with diabetic ketoacidosis and an estimated 10% uid decit. The initial 14 hours of the resuscitation were accomplished via the IO route and produced clinical improvement.24 The other was a 7-month-old infant who had dehydration caused by vomiting and was successfully resuscitated with NS through a spinal needle inserted into the tibia.25 In both cases, initial attempts at IV access had been unsuccessful. Several articles addressed IO use in neonates. In 1 study, 27 neonates in the ICU who had unsuccessful attempts at routine means of access had IO needles inserted for resuscitation. All patients had an IO catheter inserted in less than 2 minutes and were successfully initially resuscitated through the IO route with volume expanders and medications. However, 44% of the patients later died of underlying disease.26 Two case reports described successful use of IO infusions in neonates: 1 in a 38-day-old preterm infant who weighed 800 g and another in a 10-day-old infant with dehydration and poor feeding secondary to coarctation

of the aorta.27,28 In both cases, the infants were successfully resuscitated from circulatory collapse through the IO route. Few complications were noted from these studies. The authors of a study of 27 neonates reported 3 cases in which the IO catheter dislodged, 1 case of subcutaneous necrosis, and 1 case of hematoma formation.26 Two case reports of compartment syndrome after IO infusion were found: 1 in a 3-year-old who had bilateral IO infusions and 1 in an 11-month-old who had an IO infusion for 53 hours.29,30 Both patients underwent fasciotomies and recovered; 1 had minimal residual decit.30 Local swelling in 1 patient from another case report, which resolved without longterm complications.31 No recent reports of infection were identied in our literature search. In a 1946 series of 495 patients, 5 (1%) developed osteomyelitis.32 The protocols followed at that time included using iodine and hand-washing but not gloves. IP Rehydration Seven articles on IP hydration met the inclusion criteria. All were case series that ranged in size from 4 to 96 patients. All of them were conducted in developing countries, and only 1 was conducted after 1975. Reports from 6 studies noted benets with IP uids, although none of the studies had comparison groups. Patients included ranged from mildly to severely dehydrated. IP rehydration produced better results in patients who were moderately rather than severely dehydrated. The studies on IP rehydration are summarized in Table 2. Adverse effects from IP infusion were uncommon in these studies. Dyspnea was noted in 4 patients (2 of 56 children [4%] in 1 case series and 2 of 91 children [2%] in another).36,37 In 1 study, subcutaneous uid inltration was seen in 1 pediatric patient, who

REVIEW ARTICLES

TABLE 2 Articles on Intraperitoneal Rehydration

Author (Year) Agusto-Odutola (1971)
33

Population 15 patients; 323 mo old; moderate dehydration

Study Protocol 55 mL/kg IP bolus of NS with potassium

End Points Clinical rehydration; serum electrolytes; hemoglobin; PCV; urea; plasma osmolarity

Results 14 patients (93%) improved and were discharged within 24 h; 1 patient died; postmortem examination showed no complications from IP infusion; Improvements in electrolyte levels were seen by 2 h after infusion Rate of absorption was very good in 15 cases, good in 34, fair in 15, and poor in 6; remainder of patients were not accounted for; the total number of deaths was not listed; 20 postmortem examinations were performed IP-only group: all 4 children were initially normotensive and all were successfully treated. IV-followed-by-IP group: children had lower blood pressures at admission to the study and 19 (86%) of the children were successfully treated. 89.3% were fully rehydrated after 8 h; 3.6% had no response to IP uids and required IV rehydration; Signicant increase in weight and decrease in diarrhea was seen 8 h after initial infusion 14 patients with severe dehydration were excluded from the analysis after they deteriorated and needed IV uids; 74.7% of the remaining children were able to be discharged within 24 h; 11 patients died, 7 were moderately or severely dehydrated, 4 had pneumonia and malnutrition 13 children improved (93%); 1 died within 24 h

Carter34 (1953)

96 children; ages not given; patients in whom death appeared inevitable

IP bolus of D2.5NS; volume was not standardized

Rate of absorption (which was assumed to be proportional to clinical hydration state and physical examination)

Mahalanabis et al35 (1970)

26 patients; 7 mo to 6 y old; moderate-to-severe dehydration

Noerasid et al36 (1975)

56 children; 218 mo old; moderate dehydration

4 patients given only IP uids: 70100 mL/kg IP lactated Ringers solution bolus; 22 patients given IV uids for 13 h, then 5070 mL/kg IP lactated Ringers solution bolus 40 mL/kg IP bolus of 2 parts NS, one part D5 at 0 and 4 h;

Clinical state; weight; plasma specic gravity; pH; CO2

Ransome-Kuti et al37 (1969)

105 children; 91 included in the analysis; 3 mo to 4 y old; mild-to-severe dehydration

11.4 mL/kg bolus of IP uids in 1 of 5 compositions

Weight; vital signs; degree of dehydration; frequency of vomiting and diarrhea; urine output; leukocyte count; serum electrolyte levels 6-h clinical observation; hemoglobin; osmolarity; serum electrolyte levels

VanRooyen et al38 (1995)

14 children; 4 mo to 15 y old; signicant dehydration; patients for whom oral therapy and IV placement failed

80 mL/kg IP bolus of 12NS

Wenzel and Phillips39 (1971)

4 children; 510 y old; moderate dehydration

Bolus amount not specied

Clinical improvement, dened as: ability to tolerate oral uids, improved level of consciousness, absence of tachycardia and hypotension, improved skin turgor, moist mucous membranes State of hydration; stool output; urinary output

No child had improvement in clinical dehydration or stool or urine output after 4 h; all were given IV or oral rehydration and recovered; did not specify if other children were successfully treated

Denitions on the degree of dehydration varied between studies: Agusto-Odutola33 and Ransome-Kuti et al37 both used 1952 criteria from the Medical Research Council that dened moderate dehydration as a 5% to 10% decrease in body weight associated with restlessness, decreased skin turgor, cool skin, sunken eyes and anterior fontanelle, dry mucous membranes, and a pulse of 160 to 180 beats/minute. Severe dehydration was dened as 10% weight loss, with semi-coma, worsened signs of the above, and a pulse of 180 beats per minute.33,37 VanRooyen et al38 dened signicant dehydration as decreased level of consciousness, dry mucous membranes, poor skin turgor or tenting, and tachycardia or hypotension. Noerasid et al36, Wenzel and Phillips,39 and Mahalanabis et al35 did not provide criteria for their dehydration classications, although Wenzel and Phillips noted that all of their patients had tachycardia and poor skin turgor. Carter34 did not discuss the degree of dehydration but included patients in whom death appeared inevitable. PCV indicates packed cell volume; D5, 5% dextrose; D2.5NS, 2.5% dextrose NS; 12NS, 0.5% NS.

recovered without incident (ref 38 and M. J. VanRooyen, MD, MPH, personal e-mail communication, March 18, 2010). Meteorism was seen in 9 of 56
PEDIATRICS Volume 127, Number 3, March 2011

patients (16%) in 1 case series.36 The authors of this study also noted a slight increases in average leukocyte count (from 8.4 to 11.7) and tempera-

ture (from 37.7 to 38.2C) after infusion. Of 20 postmortem examinations performed after IP transfusion, there was 1 child with peritonitis that was
e753

found on autopsy. However, it was attributed to dysentery causing multiple full-thickness ulcerations of the distal ileum with one site of perforation 2 inches from the peritoneal infusion site.34 Subcutaneous Rehydration (Hypodermoclysis) Three studies met the inclusion criteria for subcutaneous rehydration. One was a series of 51 patients with mild or moderate dehydration treated with subcutaneous uid infusion assisted by a hyaluronidase enzyme. Patients were given an initial bolus of 20 mL/kg isotonic uid over 1 hour, followed by further uid if needed. Forty-three patients (84.3%) were rehydrated through subcutaneous uid in the emergency department. An additional 5 patients (10%) were reported as successfully rehydrated (judged to have been clinically rehydrated primarily through the subcutaneous route40). It is unclear if they were given any other forms of rehydration. The other 2 publications about subcutaneous rehydration date back to the 1960s. One was a case series of 4 patients with signicant metabolic abnormalities who were transferred to a pediatric center after failing to improve with subcutaneous uids. All 4 patients later improved with IV uids, and the author concluded that the subcutaneous uids were ineffective and worsened the symptoms.41 However, the volume of subcutaneous uid given to each patient was not indicated, and the author provided no reasons for postulating that subcutaneous uid worsened the patients conditions rather than the natural disease course. The other publication was a case report of a 5-month-old child with pneumonia who was given an unspecied amount of subcutaneous NS with 5% dextrose. Three days later the patient developed oliguria and edema,
e754 ROUHANI et al

which were attributed to the subcutaneous infusion.42 In the case series of 51 patients, 1 serious adverse event (cellulitis) was reported. Most patients had pain on the infusion of hyaluronidase, but twothirds of them had no pain with uid infusion.40 Proctoclysis There were no studies on proctoclysis that met the inclusion criteria. Only 3 studies on proctoclysis were identied in our literature search: 2 studies were of adults, and 1 failed to meet our study inclusion criteria because it was opinion-based rather than empirically driven.

DISCUSSION
Dehydration threatens the lives of at least 2 million children annually.1 The effectiveness of oral and IV rehydration has been well established, although at times neither is available, effective, or possible. In this review we examined the data on alternative methods of rehydration. NG rehydration for moderate and severe dehydration is supported by the results of several randomized controlled trials. The data have suggested that NG hydration is safe and effective. Despite this fact, it is not as widely used as it could be.43 Larger studies may be needed to determine if the benets of oral rehydration can be replicated with NG rehydration.44 Furthermore, although not one of our primary areas of interest, limited data suggest that compared with IV rehydration, NG rehydration results in shorter hospital stays and cost savings of as much as $115 per case in a developed country.2,15 It is important to note that, unlike IV infusion, NG infusion does not require sterile uids but can rely on standard ORS, which makes it a more accessible technique in low-resource environments.

With the exception of NG rehydration, the quality of evidence behind other methods of hydration is poor. The quality of data on IO infusions was not as robust as expected, given its widespread use as a rescue technique for intravascular access. Our literature search identied only 1 randomized trial of IO versus IV access. The results of this study and other published reports support IO infusions as effective and easy to obtain. In addition, multiple resuscitation guidelines recommend IO access in cases of shock when IV access cannot be obtained quickly. Our review has demonstrated that IO access in infants can be achieved with needles that are not specically designed for IO cannulation, such as 18gauge buttery needles19 or spinal needles.25 The quality of evidence on IP rehydration is not as robust as would be desired. Multiple case series have shown that IP hydration is effective in moderately dehydrated patients, although the results in severely dehydrated populations have been mixed. In addition, there have been no comparative studies on IP rehydration. Therefore, until further evidence is obtained, IP rehydration can only be recommended when oral, NG, IV, and IO rehydration either fail or are not available. The quality of evidence on subcutaneous rehydration is also poor. The authors of 2 older case reports warned against subcutaneous rehydration but without clear justication. Only 1 case series supported its use. However, this case series was small and designed to test a specic medication as an adjunct. As such, its results may not be generalizable. Literature on rehydration of adults has suggested that subcutaneous rehydration may be administered without hyaluronidase, although the effect on absorption is unclear.45 Although data on subcutaneous rehydration in the adult palliative

REVIEW ARTICLES

care literature suggest some benet, more research in the pediatric population is required before its use can be endorsed.46,47 Serious complications were uncommon in all of the methods we analyzed. Infection, which is a theoretical complication of all methods of rehydration, was infrequent. One case of cellulitis was reported after subcutaneous infusion.40 Low rates of osteomyelitis (1%) were reported from a 1946 series of IO infusions,32 and pooled data from adult and pediatric patients from 1942 to 1977 revealed osteomyelitis rates of 0.6%.48 More recent data from children were not available, but infection rates would likely be lower given improvements in sterile technique. The 1 case of peritonitis in a patient with IP infusion was not felt to be attributable to the infusion but to the underlying disease. Two other serious events were identied: a possible aspiration in a patient who was receiving NG rehydration and 2 cases of compartment syndrome in patients who were receiving IO infusions. Although previous studies have shown that oral rehydration has fewer complications than IV rehydration,44 we did not analyze the frequency of these adverse events in our review. As such, it is not possible to compare the frequency of complications with NG, IO, IP, or subcutaneous rehydration to those of IV or oral treatment. Overall, the data support a role for NG rehydration as an alternative to IV rehydration when patients are unable to be rehydrated orally. IV access can be used when NG rehydration fails, and IO access can be used if an IV line is not quickly obtainable. It is important to note that specialized techniques for intravascular access, including central vein catheterization and neonatal umbilical vein catheterization, were not reviewed here but remain available to

the clinician. If none of the abovementioned techniques are successful or safely available, IP rehydration and subcutaneous rehydration can be used as life-saving measures, although the data supporting their use are limited. The articles summarized here include data from resource-poor and resourcerich environments. As such, the results are applicable to either setting. However, they are likely to be particularly relevant to clinicians who work in resource-limited settings, where limited supplies necessitate exible treatment plans. In these situations, supplies such as sterile uids or tubing may be limited, which makes techniques such as NG rehydration with ORS more accessible. In addition, NG rehydration can be accomplished by parents using a syringe, which frees up limited nursing resources. In other situations, training or experience with IV cannulation may be limited. Methods such as IO, subcutaneous, or IP rehydration may represent life-saving alternatives if IV access cannot be obtained. The results of our review are limited by several factors. Our search terms, although broad, focused on hydration and may have missed patients in shock, severe trauma, or cardiac arrest. We also limited our data to human pediatric studies and required an outcome measure or complication be documented. In particular, the limited number of studies on IO infusions may have been a result of these limitations, because the IO literature may focus more on emergent resuscitation. Our data are also limited by the different inclusion criteria, protocols, and primary end points in each study. Even the denitions of degree of dehydration varied from article to article. These factors made it difcult to group studies or perform any quantitative

analysis. In addition, the sample size for most of the studies we identied was small. Even for the randomized controlled trials that compared IV to NG rehydration, sample sizes were less than 100 patients. This number may have been too small to detect a small difference in effectiveness or adverse events between techniques. The largest studies identied had no comparison groups. Finally, our review focused on the effectiveness of these techniques, but we did not consider the paucity of data on the relative discomfort experienced by patients undergoing each procedure.

CONCLUSIONS
Dehydration is a common presenting condition of children worldwide and plays a signicant role in many of the leading causes of mortality in children younger than 5 years. Various methods of pediatric rehydration are available to the clinician, each with its unique utility and limitations. In settings with limited material and human resources, it is helpful to have a clear understanding of alternative rehydration methods. Despite limited data, the results of this review provide an evidence-based approach to the hydration of a child. Oral rehydration remains the preferred route for rehydration. When oral rehydration is not possible, our review suggests that NG and IV rehydration have similar outcomes. Each method may have its unique role and advantages in different circumstances. At this time, IO rehydration is the next-best method supported by the existing literature. Additional evidence is still needed regarding IP, subcutaneous rehydration, and proctoclysis, but clinicians who work in resource-limited settings may wish to also consider these alternative methods when all other options fail or are not feasible.

PEDIATRICS Volume 127, Number 3, March 2011

e755

REFERENCES
1. United Nations Childrens Fund. State of the words children. Available at: www.unicef. org/sowc08. Accessed March 28, 2010 2. Spandorfer P, Akessandrini EA, Joffe MD, Localio R, Shaw KN. Oral versus Intravenous rehydration of moderately dehydrated children: a randomized, controlled trial. Pediatrics. 2005;115(2):295301 3. American Academy of Pediatrics, Provisional Committee on Quality Improvement, Subcommittee on Acute Gastroenteritis. Practice parameter: the management of acute gastroenteritis in young children. Pediatrics. 1996;97(3):424 436 4. Gremse DA. Effectiveness of nasogastric rehydration in hospitalized children with acute diarrhea. J Pediatr Gastroenterol Nutr. 1995;21(2):145148 5. Hidayat S, Srie Enggar KD, Pardede N, Ismail R. Nasogastric drip rehydration therapy in acute diarrhea with severe dehydration. Paediatr Indones. 1988;28(3 4):79 84 6. Nager AL, Wang VJ. Comparison of nasogastric and intravenous methods of rehydration in pediatric patients with acute dehydration. Pediatrics. 2002;109(4):566 572 7. Varavithya W, Posayanond P, Tontisirin K, Chernjitra L, Kashemsant C. Oral hydration in infantile diarrhoea. Southeast Asian J Trop Med Public Health. 1978;9(3):414 419 8. Pignatelli S, Simpor J, Ruggieri M, Musumeci S. Effectiveness of forced rehydration and early re-feeding in the treatment of acute diarrhoea in a tropical area [in Italian]. Minerva Pediatr . 2000;52(7 8): 357366 9. Pignatelli S, Musumeci S. Comparison of three oral rehydration strategies in the treatment of acute diarrhea in a tropical country. Curr Ther Res. 2003;64(3):189 202 10. Green SD. Treatment of moderate and severe dehydration by nasogastric drip. Trop Doct. 1987;17(2):86 88 11. Seriki O, Olambiwonnu NO. Management of moderately severe dehydration by continuous intra-gastric infusion of electrolyte solution. Trop Doctor. 1978;8(3):134 136 12. Sammartino L, James D, Goutzamanis J, Lines D. Nasogastric rehydration does have a role in acute paediatric bronchiolitis. J Paediatr Child Health. 2002;38(3):321322 13. Yiu WL, Smith AL, Catto-Smith AG. Nasogastric rehydration in acute gastroenteritis. J Paediatr Child Health. 2003;39(2):159 161 14. Armon K, Stephenson T, MacFaul R, Eccleston P, Werneke U. An evidence and consensus based guideline for acute diarrhoea management. Arch Dis Child. 2001;85(2): 132142 15. Duke T, Mokela D, Frank D, et al. Management of meningitis in children with oral uid restriction or intravenous uid at maintenance volumes: a randomised trial. Ann Trop Paediatr. 2002;22(2):145157 16. Perng CL, Lin HJ, Chen CJ, Lee FY, Lee SD, Lee CH. Characteristics of patients with bleeding peptic ulcer requiring emergency endoscopy and aggressive treatment. Am J Gastroenterol. 1994;89(10):18111814 17. Banerjee S, Singhi SC, Singh S, Singh M. The intraosseous route is a suitable alternative to intravenous route for uid resuscitation in severely dehydrated children. Indian Pediatr. 1994;31(12):15111520 18. Daga SR, Gosavi DV, Verma B. Intraosseous access using buttery needle. Trop Doct. 1999;29(3):142144 19. Goldstein B, Doody D, Briggs S. Emergency intraosseous infusion in severely burned children. Pediatr Emerg Care. 1990;6(3): 195197 20. Hurren JS, Dunn KW. Intraosseous infusion for burns resuscitation. Burns. 1995;21(4): 285287 21. Rutter JM. Resuscitation in a shocked infant. Br J Hosp Med. 1993;50(8):487 22. Harte FA, Chalmers PC, Walsh RF, Danker PR, Sheikh FM. Intraosseous uid administration: a parenteral alternative in pediatric resuscitation. Anesth Analg. 1987; 66(7):687 689 23. Sacchetti A, Linkenhelmer R. On the effectiveness of intraosseous infusion. J Emerg Med. 1988;6(5):433 24. Alawi KA, Morrison GC, Fraser DD, Al-Farsi S, Collier C, Kornecki A. Insulin infusion via an intraosseous needle in diabetic ketoacidosis. Anaesth Intensive Care. 2008;36(1): 110 112 25. McNamara RM, Spivey WH, Unger HD, Malone DR. Emergency applications of intraosseous infusion. J Emerg Med. 1987; 5(2):97101 26. Ellemunter H, Simma B, Trawoger R, Maurer H. Intraosseous lines in preterm and full term neonates. Arch Dis Child Fetal Neonatal Ed. 1999;80(1):F74 F75 27. Ramet J, Clybouw C, Benatar A, HachimiIdrissi S, Corne L. Successful use of an intraosseous infusion in an 800 grams preterm infant. Eur J Emerg Med. 1998;5(3): 327328 28. Martino Alba R, Ruiz Lopez MJ, Casado Flores J. Use of the intraosseous route in 35. 33. 29. resuscitation in a neonate. Intensive Care Med. 1994;20(7):529 Galpin RD, Kronick JB, Willis RB, Frewen TC. Bilateral lower extremity compartment syndromes secondary to intraosseous uid resuscitation. J Pediatr Orthop. 1991;11(6): 773776 Wright R, Reynolds S, Nachtsheim B. Compartment syndrome secondary to prolonged intraosseous infusion. Pediatr Emerg Care. 1994;10(3):157159 Evans RJ, Jewkes F, Owen G, McCabe M, Palmer D. Intraosseous infusion: a technique available for intravascular administration of drugs and uids in the child with burns. Burns. 1995;21(7):552553 Heinild S, Sondergaard T, Tudvad F. Bone marrow infusions in childhood: experiences from a thousand infusions. J Pediatr. 1947;30(4):400 412 Agusto-Odutola T. The intraperitoneal administration of physiological saline solution in the rehydration of children at the casualty department of the Lagos University teaching hospital. West Afr Med J Niger Pract. 1971;20(5):324 326 Carter FS. Intraperitoneal transfusions as a method of re-hydration in the African child. East Afr Med J. 1953;30(12):499 505 Mahalanabis D, Sack RB, Kaplan J, Jacobs B, Mondal A. Intraperitoneal uid therapy in cholera and non-cholera diarrhoea: with special emphasis on the treatment of infants and children. Bull World Health Organ. 1970;42(6):837 846 Noerasid H, Soeparto P, Rudyanto B, Sugijanto S, Hamid A, Saraswati A. Intraperitoneal uid therapy in children. Paediatr Indones. 1975;15(7 8):211218 Ransome-Kuti O, Elebute O, Agusto-Odutola T, Ransome-Kuti S. Intraperitoneal uid infusion in children with gastroenteritis. Br Med J. 1969;3(5669):500 503 VanRooyen MJ, VanRooyen JB, Sloan EP. The use of intraperitoneal infusion for the outpatient treatment of hypovolemia in Somalia. Prehosp Disaster Med. 1995;10(1):5759 Wenzel RP, Phillips RA. Intraperitoneal infusions for initial therapy of cholera. Lancet. 1971;2(7722):494 495 Allen CH, Etzwiler LS, Miller MK, et al; Increased Flow Utilizing SubcutaneouslyEnabled Pediatric Rehydration Study Collaborative Research Group. Recombinant human hyaluronidase-enabled subcutaneous pediatric rehydration. Pediatrics. 2009; 124(5). Available at: www.pediatrics.org/ cgi/content/full/124/5/e858

30.

31.

32.

34.

36.

37.

38.

39.

40.

e756

ROUHANI et al

REVIEW ARTICLES

41. Steffey JM. Hypodermoclysis in infants and children. J Iowa State Med Soc. 1963;53:393396 42. Hall BD. Complications of hypodermoclysis (re-emphasis with a case presentation). J Ky Med Assoc. 1968;66(7):626 627 43. Ozuah PO, Avner JR, Stein RE. Oral rehydration, emergency physicians, and practice parameters: a national survey. Pediatrics. 2002;109(2):259 261 44. Fonseca BK, Holdgate A, Craig JC. Enteral vs

intravenous rehydration therapy for children with gastroenteritis: a meta-analysis of randomized controlled trials. Arch Pediatr Adolesc Med. 2004;158(5):483 490 45. Constans T, Dutertre JP, Froge E. Hypodermoclysis in dehydrated elderly patients: local effects with and without hyaluronidase. J Palliat Care. 1991;7(2):10 12 46. Remington R, Hultman T. Hypodermoclysis to treat dehydration: a review of the evi-

dence. J Am Geriatr Soc . 2007;55(12): 20512055 47. Slesak G, Schnurle JW, Kinzel E, Jakob J, Dietz PK. Comparison of subcutaneous and intravenous rehydration in geriatric patients: a randomized trial. J Am Geriatr Soc. 2003;51(2):155160 48. Rosetti VA, Thompson BM, Miller J, Mateer JR, Aprahamian C. Intraosseous infusion: an alternative route of pediatric intravascular access. Ann Emerg Med. 1985;14(9):885889

PEDIATRICS Volume 127, Number 3, March 2011

e757