Critical Appraisal of The Research Pape
What is the study question?
The study question specified clearly…. “It is conducted to determine the influence of pregnancy and birth outcome with respect to maternal pre pregnancy BMI”.
What are the study factors and outcome factors?
Yes, the main study factor is pre pregnancy BMI and birth outcome of pregnant women is the outcome factor. There are other study factors like gestational age.
Who are the study participants?
Yes, study was conducted in 2010. It was conducted in West Sumatra Province of Indonesia. There is a total of 19 districts in West Sumatra. The people were selected from 3 u
an and 3 rural districts through stratified random sampling method. Pregnant women who were in the first trimester from the above selected district were selected. These women should also have their first antenatal visit between August and December 2010.
Are they likely to be representative of the population of interest?
Yes, the study consists of 1013 women but it was mainly focussed on 607 women who have complete data. Study was conducted in government run public health enters and some were recruited from private clinics. So, the study population is a representative of the population of interest.
Is the study design best way to answer the research question?
Study Design – Observational cohort study
Yes, this study is appropriate
· It is examining a rare exposure because subjects are selected by their exposure status
· Allows to assess the incidence
· Allows assessment of causality (because it gives the relationship between cause and effect i.e., BMI and birth outcome)
· Even it takes long time and it is expensive it gives strong cause effect relationship
Since the study gives strong cause effect relationship this study design is best to answer the research question.
Data collection - The data are collected using 3 questionnaires.
First questionnaire includes socio demographic factors, obstetric history, pre pregnancy medical conditions and anthropometric measures.
Second questionnaire consists of data on antenatal out comes.
Third questionnaire consists of anthropometric measures at three stages during pregnancy.
Were participants appropriately randomised?
The participants are partially randomised in the study. So, there is partial bias in the study.
Method of randomisation
Yes, the method of randomisation used here is stratified randomization
Was every one accounted for (loss to follow up)?
No, everyone is not accounted for loss to follow up. Actually, there we have no loss to follow up cases
Was loss balanced between arms
Yes, since there were no losses it was balanced
Were those lost similar to the rest of the cohort?
No, already mentioned no losses
Who was blind to treatment allocation?
No, there were no blind treatment allocation
Were there systematic differences in care?
No there were no systematic difference in care from the side of researcher. But there were changes in the way they cared themselves
Was there any e
or in assessment of outcome?
Partially, there is a small chance of e
or in the assessment of study factor and outcome factor because of the use of Asian classification and international classification in the BMI.
Was it systematic?
Yes, the study was systematic they were using standard measures.
Standard measures – Standard formula was used to determine the BMI, using self-reported pregnancy weight. It is categorised to underweight, normal weight, over weight and obese weight accordingly with the criteria of WHO.
GWG was calculated from self-reported pre pregnancy weight to the last weight measures by the midwifes in the third trimester.
LBW defined as <2500 gm at the time of delivery.
Was it differential?
Yes, they were differential they had different level of nutrition status
Is there any evidence of potential confounding?
Yes, there were evidence of potential confounder in the study
Was exposure prevalent or incident?
Yes, the exposure is an example of prevalent case not incident case
Was everyone free of the outcome at entry?
Yes, everyone is free of the outcome at entry
Were all potential confounders measured?
Yes, all potential confounders were measured in the study. A large no of confounders was used like maternal education, age parity and district(rural/u
an) for BMI comparison. There is indication of differences in baseline characteristics with different exposure levels, demonstrating that these factors are potential confounders
Were they measured accurately and without bias?
Partially, they were measured accurately without bias, because all the women with full data were selected for the study and most of the data were collected from public health centres. But bias may happen if the data are noted wrongly…that is some mistakes while collecting or typing data.
Was confounding managed by design or in analysis?
Yes, confounding was managed in the analysis. Confounding was dealt with in analysis by multivariate regression analysis.
How strong is the strength of association?
The average age difference between women’s estimated pre pregnancy weight when measured by the midwife in the first trimester was 0.92 kg.
Pre pregnancy estimated weight and actual weight in trimester 1 were highly co
elated.
GWG is higher in underweight group.
Women who were overweight had lower adjusted odds rate.
The difference in the proportion of women gaining inadequate weight using two different BMI classifications was not quite significant(p=0.037) .
Is residual confounding likely to be an issue?
Partially, in every adjustment the or move closer to the null value, so the residual confounding is partially an issue. (There is presence of confounder so we do logistic regression for the confounder whenever we enter each confounder the value moves closer to 0 (null value) so we can say the confounding effect has been change. But not completely......so that if u want u can mention residual confounding is partially affected).
Are the conclusions valid?
The author doesn’t give much idea in the conclusion, but based on what he suggests the conclusion is valid for the study