Intravenous Fluid Use in an Indonesian Community Cohort 0-18 Months of Age

Intravenous uid administration is one of the most common and important life-saving treatments in daily clinical settings, however it is not given as much attention as it should have. This study aims to evaluate the intravenous uid’s use pattern in infants from birth until 18 months of age in Indonesia. A post-hoc analysis of RV3-BB Phase IIb randomized, double-blinded, placebo-controlled trial was conducted in 1621 participants in Indonesia between January 2013 and July 2016. Any health events were documented in the trial as adverse events. Concomitant medication surveillance recorded all medications, including intravenous uids (IVF) during the 18 months of follow-up. Information included intravenous uid type, methods of administration, dosage, indication and duration. A total of 333 intravenous uid consumptions were recorded, 20 (6.0%) as uid resuscitation, 273 (81.98%) as uid maintenance, 40 (12.01%) as both maintenance and resuscitation. Isotonic uids predominated resuscitation uids (52 out of 60 [86.7%]) and maintenance uids (166 out of 313 [53.0%]). Hyponatremia (1 case) was detected among those administered with hypotonic uids (143 out 313 [45.7%] IVFs) for maintenance uid. The most common indication for intravenous uid was gastroenteritis (31.6%), with dehydration in most cases (65 [69.9%]). Modied WHO’s plan B diarrhea management was recorded in 22 [23.7%] diarrhea cases. (0.3%) uid 1 (0.3%) colloid (2 recorded hypotonic hypotonic infection used combination other used dextrose 5% alone for benign neoplasm of bronchiolitis, rhinopharyngitis, Sturge-Weber treat Hydroxyethyl was used to treat dengue used either chloride a


Introduction
Intravenous Fluids (IVF) are commonly indicated for acutely ill children who cannot receive su cient uid by enteral administration alone. This could be due to gastrointestinal disease, respiratory compromise, neurologic impairment, cardiac arrest, hypovolemic shock and other life-threatening conditions. (1)(2)(3)(4) The main indications for IVF use include : (1) Cardio-vascular resuscitation, (2) Replacement of uid and electrolyte de cits, and (3) to provide routine maintenance uid and electrolyte requirements (5) with recently added consideration to (4) Redistribution of the uid and (5) Reassessment of the IVF need during the therapy. (6) These '5 Rs' should always be considered before and during administration of IVF to patients. (6) Intravenous uid should be administered adhering to the same rigorous standards as is required for drugs,, and if there is no indication for IVF administration, then no IVF should be given. (3,4) However, in clinical practice, these standards are all to commonly not be applied consistently, and IVF are sometimes administered despite not meeting an accepted indication, and not meeting recommendations for the dose and type of IVF for that clinical scenario. While IVF administration is a routine procedure in hospitals, research efforts allocated into this topic are still scarce. (4,7,8) Complications resulting from inappropriate and incorrect IVF administration can be harmful. These range from local minor complications caused by incorrect intravenous cannulation (e.g. clotting, occlusion, leakage, in ltration, extravasation, phlebitis and infection) to systemic and lethal complications of incorrect intravenous uid administration (e.g. hypo-or hypernatremia). (3,4,9) Approximately 1 in 5 adult patients are reported to suffer harm from the complications of inappropriate IVF administration. (6) In order to tackle this issue we have to rst understand the current practice of IVF administration in different settings.
In a WHO report published in 2011, an injection rate ranging from 0-7.8% across 7 public facilities and 5 private facilities was reported in Indonesia, based on data provided from the Indonesian Ministry of Health. However, these data did not include injection and infusion use in any emergency department. There are no data speci cally regarding the use of IVF across hospital inpatient and outpatient services in Indonesia. (10) To critically evaluate the use of IVF in Indonesia, including the appropriateness, safety and administration practices studies on are required. As part of a post-hoc analysis for RV3-BB vaccine trial population, along with the previous antimicrobial and non-antimicrobial studies, this study focused on IVF administration patterns in the trial participants. (11,12) This study may help Indonesian and International authorities to assess and develop strategies in IVF administration policies.

Study design
The RV3-BB rotavirus vaccine phase IIb randomized, double-blinded, placebo-controlled trial (Australian New Zealand Clinical Trials Registry number ACTRN12612001282875, registered date: 12/12/2012; the protocol is available at NEJM.org) was primarily done to assess the e cacy of the vaccine against severe rotavirus gastroenteritis on children up to 18 months old. Trial participants were divided into three randomised groups: neonatal-schedule and infantschedule vaccination, and placebo. Adverse events surveillance includes scheduled visit and weekly telephone calls by study personnel. The primary outcome from this study are episodes of severe rotavirus gastroenteritis from 2 weeks after 4 doses of investigational product. The secondary data obtained from this study are used to analyse the Intravenous Fluid use pattern on the participants during this study. The full design of this study will not be explained here as it has already been clari ed in the main study. (13) This study aims to evaluate the intravenous uid's use pattern in infants from birth until 18 months of age in Indonesia.

Participants
The main study, which was done in two districts of Yogyakarta and Central Java Province, Indonesia, from January 2013 to July 2016, includes 1649 participants, of which 1621 participants' data are used in this study. There are 49 participants who were lost to follow up, however we still include their data in our study since some of them used at least one IVF and other medication. Figure 1 shows the pathway of how the participants were recruited. The data were obtained from 23 Primary Health Centres (PHCs) and 2 hospitals in the Sleman district (Yogyakarta Province) and the Klaten district (Central Java Province), representing the urban and rural area respectively. Healthy and full-term babies who weighed between 2500 and 4000 grams were included in this study, following written informed consent. Intervention, randomisation and blinding The participants were divided into three groups: neonatal-schedule vaccine group, infant-schedule vaccine group, and placebo group. The participants received 4 doses, which consisted of 3 vaccine doses and 1 placebo dose, with exception for the placebo group who received 4 placebo doses. The schedule for the vaccinations was as follows: dose 1 at 0-5 days of age, dose 2 at 8-10 weeks of age, dose 3 at 14-16 weeks of age, and dose 4 at 18-20 weeks of age. The neonatal-schedule group received the placebo at the 4th dose, while the infant-schedule group received the placebo at the 1st dose.
The group intervention assignment was randomized with a computer code with a block size of 6 strati ed by district. The trial assignment information was only available to the vaccine or placebo dispensing pharmacist at central pharmacy in each province, while both investigators and participants were blinded.

Sample size
To provide a power of 80%, one-sided alpha of 0.1 and allow 10% dropout for the primary outcome of the assumed 3% severe gastroenteritis rotavirus, 549 participants per group were needed. In sub-study A, to compare the cumulative vaccine take, 94 participants per group were needed, meanwhile in sub-study B, to describe the proportion of infants with a positive seroprotective response against polio strains 1-3, 111 participants per group were needed. Since the sample size for the primary outcome is the largest, it is used as the study's sample size. The complete sample size calculation is described in the main study publication. (13) Intravenous Fluid All the participants who were included in this study were questioned weekly by phone call for any illness and medications (including intravenous uid) that were administered to them during the 18-month period study. If they got sick, then attended the Primary health care clinic, or were hospitalized, then the data concerning intravenous uids given to them are extracted from their medical records into electronic case report forms. In this study we did not give any treatment recommendation, including types of IVFs used and their indication. Treatment decisions were completely entrusted to the attending clinicians and routine local practices. Data obtained include intravenous uid type, methods of administration, dosage, indication, start date and end date.
Based on WHO ATC codes (14) the IVFs are classi ed as (1) combinations of electrolytes, (2) combinations of electrolytes with carbohydrates, (3) carbohydrates, (4) electrolytes, (5) other blood products, and (6) blood substitutes and plasma protein fractions. All of these types of uids, except for Other blood products and blood substitutes and plasma protein fractions, are what we call crystalloids. Other blood products and blood substitutes and plasma protein fractions can be regarded as colloids.
All illnesses or symptoms that occurred after the rst dose of vaccine or placebo within the 18 months observation period were recorded as Adverse event (AE). AE were later classi ed with Medical Dictionary for Regulatory Activities (MedDRA) and subsequently narrowed down to 13 disease classes according to the affected organ systems. For diarrhea cases, dehydration level are based on WHO's Integrated Management of Childhood Illnesses (IMCI) guideline. (15) Ethical considerations

Statistical analysis
Statistical calculations were performed by using SPSS version 25. For descriptive data, the results were presented as mean, median, frequency and percentages. Relationship between gender and vaccination groups with IV uid use was assessed with Pearson chi-square. Relationships between the number of AEs and IVF Therapy use was assessed with independent samples t-test. Kolmogorov-Smirnoff test were used before hand to determine the normality for AEs for each participant.

Baseline characteristics
Out of 1621 participants enrolled in this study, 251 participants (15.5%) received Intravenous Fluid (IVF) at least once in the study period. Baseline characteristics of the participants are shown in Table 1 below. There are no signi cant differences in IVF use between gender groups and vaccination groups. Adverse events (AE) recorded per individual are signi cantly associated with IVF therapy (p < 0.05).

Intravenous Fluid administration pattern
The histogram in Fig. 2 depicts the age of participants at the start of intravenous uid therapy. Notice that the IVF frequency decreases during the rst 3 months of life, then peaks at the 9th month of life.
As seen in Table 2

Types and indications of IVF use
The types of IVF used in participants in this study are presented in Table 3. The characteristics and composition of the IVFs administered are presented in Table 4. The majority of the IVF administered in this study were crystalloid solutions with isotonic solutions being the mostly used (54.1%) ( Table 5). The IVF administration duration, as shown in Table 5, was not signi cantly different between the group that received isotonic IVF solution when compared to the group that received a hypotonic IVF solution (p > 0.05), but participants who receive hypotonic uids are younger than those who receive isotonic uids (p < 0.001). Intravenous Fluids which were used for AEs requiring resuscitation (n = 60) consist of 52 (86.7%) isotonic uids, 4 (6.7%) blood products, 2 (3.3%) colloids and     * Data are shown as no. of diarrhea event (% of total) †Custom regimen is described as a customized uid management but cannot be categorized into any plan based on IMCI guideline. ‡ Unknown regimen is described as unavailable data on speci ed uid management from the case report form.

Discussion
There are no previous studies that report the rate of IVF treatment in Indonesian community. Studies of the use of IVF are still limited which urge the need of further studies of IVF use in children. The age of the event onset requiring IVFs (Fig. 2) suggested that intravenous uid use was decreasing starting from 3 months old, before starting to to raise until the peak at 9 months old. In this study, participants who received IVF in the neonatal period were associated with gastrointestinal disorders and infectious diseases (predominantly neonatal sepsis). The major disease occurring at rst 2 months of age was respiratory system disorders, while the older ages were predominantly gastrointestinal disorders and respiratory disorders.
IVFs in this study was almost exclusively administered by peripheral IV catheterization. About 55/294 (18.7%) events with IV cathether in this study spent more than 5 days (but less than 15 days. The Michigan Appropriateness Guide for Intravenous Catheters (MAGIC) mentions 7 methods of IVF administration.
The most common method, peripheral IV catheterization, is recommended for anticipated IV administration duration of less than 6 days. Beyond that period, other methods such as ultrasonography guided peripheral intravenous catheterization are preferred. (20) Estimating the duration of IV administration, however, is di cult. Ultrasound-guided IV cannulation is more commonly indicated for di cult IV access, such as in obese patients. (21) Absence of other IV catheterization technique (other than peripheral IV catheterization and IO catheterization) in the participants shows the need to explore this topic more, as it might help to improve IV practices in Indonesia.
More than half of total AE indications, based on our study, were given IVF and other concomitant drugs. Peripheral intravenous catheterization is indicated for intravenous (IV) drug administration, IV hydration, transfusion, surgery, emergency care, and in other situations that requires direct IV access. for ongoing body uid loss; plan B, intended for some dehydration, which require ORS for uid replacement and ongoing loss, except for participants with profuse vomiting, who can be treated with IVFs; and plan C, intended for severe dehydration, which require uid replacement via intravenous or Nasogastric Tube and ORS to replace ongoing loss. (2,15,25) This nding shows that only 1 (1%) IVF was used for the right indication and with the right regimen in diarrhea events.
In this study, isotonic uid was the most used uid, but a high percentage of hypotonic uid was still being used. Hypotonic uids tend to be given to younger participants; this shows that some IVF practices were still based on old theory made by Holliday-Segar. (26) The most frequent IVFs used for maintenance is isotonic uid, but a large proportion of hypotonic uids are also used for maintenance uid. The Indonesian Pediatrics Society's handbook mentioned how much uid should be administered based on the patient's weight and they also mentioned to use Dextrose 5% in 0.45% NaCl solution as the uid of choice for maintenance of Diabetic Ketoacidosis case; however, there is no recommendation of IVF of choice in general. (25) Long standing WHO recommendations stated to use either isotonic or hypotonic IVFs for maintenance IVFs (2) but present studies recommends only isotonic uids for maintenance (except for neonates) because it is less likely to cause adverse events in children. (1,19,24) A study in England (2015), showed that 118 (59%), 63 (32%) and 19 (10%) pediatric patients receive hypotonic, isotonic and hypertonic maintenance respectively. (27) Compared to their nding, maintenance IVF practices in the study showed better adherence to isotonic uids as maintenance uid. Even then, we cannot disregard the fact that a large proportion of maintenance uid in this study were hypotonic uids.
In our study, 1 case of hyponatremia was detected after a hypotonic uid administration indicated for pneumonia, on the contrary there is no hyponatremia detected in those who were treated with isotonic uid. This shows that uid administration practices can still be optimized to reduce the risk of complications by simply selecting the right choice of uids.
The most used resuscitation uid is isotonic uids, with only a small fraction of hypotonic uids being used. This nding shows that clinical practice of uid resuscitation during the duration of the study is good and almost all of clinicians comply to Indonesian Pediatric Society handbook guideline to use isotonic uid for resuscitation. (25) Recommendations for resuscitative uids is adamant on using isotonic uids (1,2,19,24), except perhaps in case of malnutrition.
(2) Comparisons of different recommendations for maintenance and resuscitative uids from different sources are described at Table 8.  Some intravenous uid choice for children with diarrhea used hypotonic uid as maintenance in this study, which is not optimal and shows that IVF practices have not been in compliance with the latest recommendation.
In this study, blood transfusion was performed in 2 participants with 9.3-9.6 g/dL for pretransfusion hemoglobin (Hb) level and 13.3-14.1 g/dL for posttransfusion Hb level. Latest recommendations state that the more restrictive Hb threshold (7 g/dL) for transfusion reduces blood use and does not result in worse outcome in pediatric intensive care unit. (28) The target Hb concentration recommended is around 8.5-9.5 g/dL. (29) The higher threshold might be justi able when indicated for the participant with intracranial hemorrhage, because the participant is hemodynamically unstable, but questionable for the participant with cholestasis. Implementation of more restrictive Hb threshold might give more bene t than harm.
Colloid use in this study was Hydroxyethyl starch (HES) and albumin indicated for dengue hemorrhagic fever and cholestasis, respectively. The participant treated with HES was suffering from a fatal dengue shock syndrome. After resuscitation with Ringer's Lactate, she was given HES, following WHO's treatment algorithm (30). Albumin is not only intended for volume expansion. In this study it was intended to correct hypoalbuminemia at a concentration of 2 g/dL in a case of cholestasis. Although Indonesian guideline recommended to correct albumin in case of hypoalbuminemia in nephrotic syndrome, there is no recommendation in cholestasis. (25) furthermore, this effort to correct albumin without other primary indication such as hypovolemia or major surgery might be unnecessary. (31,32) The study re ects the patterns of IVF administration use in infants and young children in Yogyakarta and Central Java province, and as far as we know, IVFs were administered improperly in the community, and some choice of intravenous uid used for the participants in this study is not optimal. Practitioners' knowledge and skills towards intravenous uid management in children plays a major role in determining the rationality of the treatment, and Mahapatra et al.
found that practitioners who live in poor and resource-scarce area have a tendency to be less competent than those who lived in more developed area, which might results in more irrational IVF administration. (33) Rationality of IVF administration in this study can only be assessed in the way IVF resuscitation were given to participants with diarrhea based on their dehydration level and the result shows more questions rather than answers. Maintenance uid administration was needed in surgical patients (congenital megacolon, hernia, phimosis, etc.). However, for the rest of those who receive maintenance uid without accompanying IV drugs and speci c indications (e.g. dengue fever and febrile convulsion) which necessitate IVF treatment, its rationale requires further assessment on the patient's ability to intake uids orally as stated by the NICE guideline. (6) Conclusion This study provides the information of intravenous uid use pattern in 0-18 months children cohort in Indonesia. Some intravenous uids were found to be unnecessary and not optimal, exposing the patients to the avoidable risks. These data justify the further study to evaluate intravenous uid use in children and the needs of continuing training on intravenous uid management among healthcare providers.

Declarations
Ethics approval and consent to participate

Consent for publication
Not applicable Availability of data and materials The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
N.S.B. is employee of PT Bio Farma who provided funds to support this study. All other authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.

Funding
This trial was funded by the Bill and Melinda Gates Foundation, the National Health and Medical Research Council, and PT Bio Farma.
Authors' contributions JAT, CDS, KPJ, and EW made substantial contributions to designing the study, analyzing the data, interpreting the results, writing and revising the manuscript.