Treatments for infant colic |
||
|
|
A systematic review of treatments with NNTs in the abstract [1] seems like a godsend, especially when some of those NNTs are low. But beware reading abstracts, because behind the abstract we learn how flimsy are the conclusions for those NNTs. A frustration with the evidence-based approach can be that it tells us just how little we know. Defining the research agenda is wonderful for academics, but we have a screaming child and we want the answer NOW! Bandolier has misgivings about evidence when there's actually precious little of it. But infant colic affects up to 150,000 of the children born in the UK every year, and for such a common condition, on balance, a brief reprise of the evidence available probably makes sense.
Search
The review sought randomised trials of infant colic in English from MEDLINE, the Cochrane Library, books and review articles.
Results
There were 23 comparisons in 22 reports, 12 of which were considered adequately to describe double-blinding, and nine of which adequately defined cases using the Wessel definition of colic. This is unexplained paroxysmal bouts of fussing and crying that lasted more than three hours a day, for more than three days a week, and for more than three weeks. Seven studies examined pharmaceutical interventions, eight dietary interventions, four behavioural interventions and three naturopathic interventions. The trials were small. Three comparisons had fewer than 25 infants, 12 between 26 and 50, five between 51 and 100 and three between 101 and 166.
The results are shown in the Table. While occasional studies demonstrated some effect, most are likely to be subject to bias. Sucrose studies were of reasonable standard, if small, but showed only brief effect. Dicyclomine studies were of variable quality, but potentially severe adverse effects, including apnoea, preclude its use. Hypoallergenic diets for mothers and infants in studies of mixed quality may have small effects, but their reporting is confusing, and confounded by small numbers, as most of these trials.
Table: Trials of interventions for infantile colic
| Treatment | Number of trials | Number of infants | Comment |
| Simethicone | 3 | 272 | Three trials with adequate double blinding showed no evidence for efficacy |
| Dicyclomine | 3 | 134 | Dicyclomine better than placebo in three trials, one adequately double blind. Adverse effects, some serious reported in infants and drug contraindicated in infants less than 6 months |
| Soy formula | 2 | 158 | One study showed good improvement, while larger of the two did not permit analysis. |
| Increased carrying | 2 | 94 | No effect |
| Hypoallergenic formula | 2 | 72 | Some indication that hypoallergenic formula has a beneficial effect in two studies |
| Sucrose | 2 | 72 | Sucrose appears to be effective only for a short while |
| Lactase enzymes | 2 | 44 | No difference between lactase enzymes and placebo |
| Hypoallergenic diet | 1 | 115 | Complicated because breast feeding mothers and bottle-fed infants included in trial. May be a small reduction of daily colic symptoms of 25%, but complicated design and inconsistent result reporting |
| Herbal tea | 1 | 68 | May be slight effect in single study, but sensible parents unlikely to try this in small infants |
| Fibre enriched formula | 1 | 54 | No effect |
| Decreased stimulation | 1 | 42 | Bare significance in trial with potential for much bias |
| Methylscopolamine | 1 | 40 | No benefit of treatment and more adverse effects |
| Dairy elimination diet | 1 | 40 | No benefit of treatment |
| Car ride simulator | 1 | 32 | No effect |
| Parent training | 1 | 14 | No effect in tiny flawed trial |