Original Articles Indian Pediatrics 2000;37: 952-960 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Controlled
Trial of Hypo-Osmalar Versus World Health Organization Oral |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Manuscript received: May 10, 1999;
Initial review completed: June 30, 1999;
According to the ESPGAN working group (1) an oral rehydration solution for the European children can safely contain glucose between 74-111 mmol/liter and a sodium concentration of 60 mmol/liter. Due to the passage of low sodium diarrhea in European children, a low osmolarity (200-250 mosmol/liter) ORS will ensure optimal water absorption with adequate replacement of sodium, potassium and base. Further, open(2) and double blind trial(3) have demonstrated the efficacy of hypotonic ORS in acute diarrhea. Experimental studies(4,5) on a human model of cholera suggest that improved rate of rehydration is possible by use of hypotonic ORS with mid range sodium concentrations. Thillainayagam et al.(6) have shown more net absorption of water in normal and secretory rat intestine with hypotonic ORS. In the early nineties, when EI Mougi et al.(7) found lower serum sodium, stool output and duration of diarrhea with the diluted (ORS packet dissolved in 1.5 liters of water) ORS solution, the developing world faced the key question Is low osmolarity ORS more efficacious than the standard WHO-ORS solution?. The International Study Group on reduced osmolarity ORS solutions in their multicentric trial(8) has also reported beneficial effects on the clinical course of acute non-cholera diarrhea. Others validated their results in child-hood non-cholera diarrhea(9,10), in rotavirus negative childhood diarrhea(11) and adult cholera(12,13). This study was designed to compare the efficacy and the safety of the hypo-osmolar ORS (H-ORS) with the standard WHO-ORS, in childhood cholera and non-cholera diarrhea.
All children with acute (of <4 days duration) diarrhea with dehydration were included in the study, if they met one of the following criterion: (i) non-cholera diarrhea, with age between 3 months to 5 years; and (ii) all children above 3 months with clinical suspi-cion of cholera. Children with clinical evidence of systemic infection, encephalopathy, electrolyte im-balance, convulsions or invasive diarrhea were excluded from the study. All cases were serially allotted the study ORS packet, containing 5 sachets, each to be formulated in 1 liter of water. Complete admis-sion characteristics of the patients were filled in a proforma and if severely dehydrated, they received 50 ml/kg of intravenous Ringer's lactate in the first hour, before inclusion in the study. The included children were advised 75 ml/kg of ORS in the next 4 hours. Subsequent advice was given as per the recommended plan A, B, and C(14). Intake output records and assessment of dehydration were done 4 hourly. Frequency of stools was recorded by the mother by tally marking. Urine output collection was done for boys for the initial 24 hours. Weight was taken at admission, at the end of rehydration and at discharge. The nutritional status was assessed as per National Center of Health Statistics (NCHS). Khichri, Dalia, curds and banana feeds were offered, once hydration improved. Breastfeeding was continued throughout. Stool for examination for motile organisms and stool culture was done for all patients. Stool for culture was collected on sterile rectal swab, stored in Careyblair's media and plated within 12 hours. A single dose of doxycycline (8 mg/kg) was administered to all patients (even to less than 1-year-old) with clinical suspicion of cholera or positive stool for motile organisms. The dose was repeated if the child vomited within half an hour of taking the drug. Serum sodium estimation was done at 24 hours. The following patients were considered as failure of study ORS and were shifted to known WHO-ORS: dehydration >72 hours; diarrhea >7 days; consumption of ORS >8 liters in <5 years age group, or >10 liters in >5 years age group; and needing intravenous fluids more than 150 ml/kg. Patients who left the study before recovery were also considered in the failure group, if they had dehydration and or frequency of stools ³10 in the last 24 hours, at the time of leaving. Recovery and discharge criteria for non cholera diarrhea was three consecutive semiformed stools or no stools for 12 hours, while that for the cholera cases was no dehydration for 8 hours or no stools for 6 hours. Indications for intravenous fluids were severe dehydration, persistent vomiting (>3h), and persistent dehydration at the end of 4 hours of oral rehydration therapy. 75 ml/kg of Ringer's lactate were given in the next 3 hours, and then the child was put back on the study ORS. WHO ORS and hypo-osmolar ORS packets were prepared in the departmental Research laboratory. The composition of hypo-osmolar ORS (NaCl 2.6 g, KCl 1.5 g, Trisodium citrate 2.9 g, Glucose 13.5 g) and WHO ORS (NaCl 3.5 g KCl 1.5 g, Trisodium citrate 2.9 g, Glucose 20 g) was different in their amounts of NaCl and glucose. Hence, the concentration of sodium, glucose, and chloride in the hypo-osmolar ORS versus WHO ORS were 75 vs 90 mmol/L, 75 vs 111 mmol/L and 65 vs 80 mmol/L, respectively. The concentrations of potassium (20 mmol/L) and citrate (10 mmol/L) were identical in the two ORS. The resultant osmolarity of the hypo-osmolar ORS was 245 mosmol/L compared to the 311 mosmol/L of WHO ORS. Since this was a double blind clinical trial, identical packets containing 5 sachets of the salts were randomized and serially given a number by a faculty colleague not involved in the study. An additional five-sachet packet was also available for each packet allotted. Analysis of the different parameters was done in the rehydration phase, maintenance phase separately and also as overall combined data. For analysis, the children were split into cholera and non-cholera, and also considered as one total group. The analysis was repeated for children with weight for height <70% and breast fed and non-breastfed children under 2 years of age. The study was planned to detect a 30% difference in the frequency and duration of diarrhea of the two ORS under study. It was calculated(15,16) that 82 children were needed in each group to detect this difference with a power of 90% and a significance level of 5%. Previous data (frequency of 4.11 ± 2.67 stools/4 hours and duration of diarrhea 36 ± 20.08 hours) from the Diarrhea Treatment and Training Unit of 70 noncholera children treated on WHO-ORS was used to determine the sample size. Statistical analysis was done with the help of SPSS (Verson 7.5). All variables with skewed distribution were log transformed. A 2-tailed Student's t-test was used to compare the groups. Chi square test was used to correlate the qualitative variables.
We enrolled 170 cases till August 1998, of which 35 were culture proven cholera. Of these 170 children, we discharged 151 after recovery. Two children recovered after rehydration phase itself. Eleven children left the study before recovery. Four of these had a frequency of ³10 in last 24 hours, at the time of leaving the study and were considered as failures. Also considered failures were eight cases that had to be removed from the study and put on a known WHO-ORS. Of these, 6 required more than the predetermined volume of ORS, and 1 had a dehydration phase of more than 72 hours, while one needed intravenous fluids more than 150ml/kg. We therefore, had a total of 12 (7%) failure cases in the study. The data collected till their stay in the study has been included in the analysis. Of the 170 children enrolled, 88 were assigned H-ORS and 82 WHO-ORS. The mean duration (p = 0.6, 95% CI 11.9-20.5) and fre-quency (p = 0.79, 95% CI 1.1-1.4) of diarrhea at admission showed no significant difference in the two groups. Percentage of children with vomiting (OR 1.06, 96% CI 0.43-2.31), with some (OR 0.89, 95% CI 0.71-1.12) or severe (OR-1.61, 95% CI 0.59-4.33) dehydration and those receiving ORS (OR 0.74, 95% CI 0.24-2.22) at admission were comparable. Of the 35 cholera cases, 19 received H-ORS and 16 WHO-ORS. In the latter group, we had 3 failures, while an additional patient left the study after rehydration; the data of the maintenance phase of these patients is not available for analysis. Frequency and duration of diarrhea, during both rehydration and main-tenance, albeit showing a decreasing trend in the H-ORS group, were comparable. Overall, only frequency was significantly decreased (p = 0.049; 95% CI 0.38-0.99) (Table I). Although, the H-ORS was consumed more in the rehydration phase as well as overall, but in the maintenance phase the mean H-ORS consumed was significantly lower (p = 0.04, 95% CI 0.44-0.98). Of the 135 non-cholera cases, 69 received WHO-ORS. On account of failure or having left the study, the maintenance phase data was available in 67 H-ORS and 62 WHO-ORS children. Mean frequency, duration and ORS consumed during rehydration phase, main-tenance phase and overall, though largely show-ing a decreasing trend in H-ORS group, was comparable (Table II). The amount of intravenous fluid received was significantly (p = 0.045, 95% CI 1.01-2.32) more in the H-ORS group. The combined data of total cases shows comparable (OR 0.28, 95% CI 0.07-1.1) number of failures in H-ORS (3 of 88) and WHO-ORS (9 of 82) groups. Of the total 170 cases, the duration of rehydration phase, maintenance phase and overall, though showing decreasing trend in the H-ORS group, were comparable between the two groups. The frequency of diarrhea was significantly lower only in the rehydration phase (p = 0.04, 95% CI 0.64-0.99) and overall (p = 0.046, 95%CI 0.70-0.99). But the amount of intravenous fluids needed was significantly (p = 0.026, 95% CI 1.06-2.14) more in the H-ORS group (Table III). There was no significant (OR 1.54, 95% CI 0.83-2.8) difference in the proportion of children who vomited in the two ORS groups. In the severely malnourished (weight for height < 70%) group, 41 received H-ORS with 40 WHO-ORS. The rehydration (p = 0.32, 95% CI 0.55-0.97), maintenance (p = 0.035, 95% CI 0.51-0.97) and overall (p = 0.011, 95% CI 0.56-0.93) frequency of stools and mainte-nance duration (p = 0.007, 95% CI 0.46-0.88) were significantly less in those receiving H-ORS (Table IV). On repeating the analysis in 82 breast-fed children under 2 years of age, there was no significant difference in the 45 H-ORS and 37 WHO-ORS group. Of the 33 under 2-years non breast-fed children, 16 and 17 were respectively in the H-ORS and WHO-ORS groups. In non-breast fed children on H-ORS, the total duration of diarrhea (23.53 ± 1.6 vs 34.8 ± 1.7) was significantly decreased ( p = 0.03, 95% CI 11.07-11.45), while the amount of intravenous fluids (180 ± 0 vs 69.6 ± 1.01) needed was significantly increased ( p = 0.02, 95% CI 109.8-112.1). One of the 16 children on H-ORS and 3 of the 17 on WHO-ORS needed intravenous fluids among nonbreastfed children. TABLE IGeometric Mean (SD) of H-ORS versus WHO ORS in Cholera
*n(WHO-ORS)
= 12, **n(H-ORS and WHO-ORS) = 12. TABLE IIGeometric Mean (SD) of H-ORS versus WHO ORS in Non-Cholera Diarrhea
*n(H-ORS) = 62, **n(H-ORS) = 41 and n (WHO-ORS) = 40. ***n(H-ORS) = 11 and n(WHO-ORS) = 9. Except +, none of the differences between the two groups were significant (p >0.05). +: p<0.05.
TABLE IIIGeometric Mean (SD) of H-ORS versus WHO ORS in Total Cases
* n (H-ORS) = 86 and n (WHO-ORS) = 74, **n (H-ORS) = 53 and n (WHO-ORS) = 52. *** n (H-ORS) = 11 and n(WHO-ORS) = 12. Except + none of the differences between the two groups were significant (p >0.05). + p <0.05.
TABLE IVGeometric Mean (SD) of H-ORS versus WHO ORS in Severe Malnutrition.
* n (H-ORS) = 22 and n (WHO-ORS) = 19, ** n (H-ORS) = 4 and n (WHO-ORS) = 7. Except +, none of the differences between the two groups were significant. + p <0.05.
Other than the reduced frequency of stools in total cases, the present study shows no beneficial effect of hyposmolar ORS in rehydration phase in cholera, non cholera or total cases. This finding corresponds with the lack of beneficial effect seen in reduced osmolarity ORS in the first 24 hours of diarrhea in other studies(10,11). Some have reported significant decrease in stool output(8,9,13), ORS intake(8,9), and number of children vomiting(9) during the first 24 hours. Consistent with other studies(10,11) we found no significant difference in the serum sodium level in the two groups. Other authors(8,9) found a significantly lowered serum sodium at 24 hours in reduced osmolarity ORS group. Mean urine output in boys at 24 hours of admission was similar in both the groups as also reported earlier(8,9,11). The studies(9,11) which discuss the main-tenance phase or the second 24 hours of diarrhea have again not shown any significant difference in the parameters between the two ORS groups. In a study done on adult cholera(12), the beneficial effects of the reduced osmolarity ORS were seen on stool volume, duration of diarrhea, and ORS intake, whereas the serum sodium was comparable. In a Bangladesh study(11), when rotavirus negative cases were analyzed, there was significant decrease in ORS consumed and stool output in the second 24 hours of the diarrhea, when using hypotonic ORS. In the present study, significantly less ORS was consumed in the cholera cases when treated with H-ORS during the maintenance phase. The same group had required more (p >0.05) ORS in the rehydradation phase. The beneficial effect of maintenance phase is lost when we analyze the overall data in the present study. The multi-centric trial(8) done on non-cholera diarrhea showed significantly more risk of receiving intravenous fluids in reduced osmolarity ORS group when Indian children were analyzed. In the present study also, the amount of intravenous fluid needed was significantly more in the H-ORS group when we analyzed the non-cholera and total cases. Interestingly, on analysis of the cholera cases, this was not so. On the other hand, the pooled data(17) from two studies on adult cholera showed that the need for intravenous fluid actually decreased significantly in hypo-osmolar ORS group. Less but insignificantly so, overall consumption of WHO-ORS in cholera in this study strengthens the status of WHO-ORS in cholera. None as against 3 children needed intravenous fluids in the H-ORS and WHO-ORS groups, respectively in cholera, which could be a chance finding considering the small numbers. However, in noncholera diarrhea despite decreas-ing trend in H-ORS consumption, significantly more amount of intravenous fluids needed by them again favours the already established WHO-ORS in the management of childhood diarrhea. The proportion of children vomiting and the failure rate was comparable in the two groups in the present study. Some workers found vomiting(3,11) and failure rate(3) significantly higher in the children treated with the standard WHO-ORS, whereas EI-Mougi et al.(10) found no difference in the failure rate with the two ORS groups. In the non-breastfed children we found contradictory finding of significantly greater need of intravenous infusion which is contrary to the multicentric trial(8) and decreased duration of diarrhea in the H-ORS group. The beneficial effect of H-ORS on the duration of diarrhea in non-breastfed children seen in this study can be compared with work done in the developed world(2) where the proportion of non breastfed children would be predominant. The small sample size of non-breastfed children in the present work where only 4 children needed intravenous fluids does not justify any inference to be drawn from these findings. The caloric intake and the weight gain were comparable, which is consistent with the earlier studies(1,6,9). We found reduced frequency (in rehydration, maintenance and over-all) and decreased duration (of maintenance phase) in severely malnourished children. These findings cannot be compared with other work since all previous studies have included children with weight for height more than 80%. The benefit could be attributed to glucose malabsorption in malnutrition, which needs to be validated. Although, isloated beneficial effects are seen with the H-ORS used in the study, they seem to be lost when we see the overall data. The benefits of the H-ORS in maintenance phase in this study have been seen earlier(7), by encouraging sips of plain water with WHO-ORS. Further, a definite increase in the amount of intravenous fluids needed is seen with the H-ORS. Hence, in this study, we conclude, that in spite of a decresing trend in various parameters of diarrhea with H-ORS, its use shows no signi-ficant benefits over the WHO-ORS in the overall analysis of cholera, non-cholera and total cases of diarrhea. There may be some role of H-ORS in malnourished children, but further studies are required to clarify that. Considering the massive financial burden involved, and the lack of clinical significance of the benefits of H-ORS, it is not currently feasible to press for a change to the reduced osmolarity formulations worldwide.
This study was possible with the material support and encouragement provided by the Department of Pediatrics under the chairmanship of Prof. M. Ashraf Malik. We acknowledge the contribution of Mr. Hasan Nawaz Waris, Senior Technical Assistant, Department of Pediatrics, in the preparation of the oral rehydration salts. Contributors: SA coordinated the study, particularly its design and interpretation and drafted the paper. She will act as the guarantor of the article. KA, MM and IS were involved in the data collection and drafting the article. Funding:
Department of Pediatrics, JNMC, AMU, Aligarh for raw material and
preparation of ORS.
|