Essay about Critical Appraisal

Introduction

This essay will critically appraise a paper investigating the relationship between moderate alcohol intake during pregnancy and risk of foetal death written by Andersen et al. (2012). This was a cohort study that used the data from a sample of 91,843 Danish mothers. It was found that 55% of the cohort abstained completely during pregnancy, whilst the rest of the cohort reported consuming alcoholic drinks during this time. The authors reported that there was a substantially increased risk of either spontaneous abortion or stillbirth in women who consumed even moderately low amounts of alcohol (2-3.5 drinks per week) before their 16th week of pregnancy. However, alcohol consumption after 16 weeks appeared not to have an effect.

The introduction of this paper is extremely short, although it does manage to summarise why the study was conducted and the aim of the research is clear. The authors highlight that previous research into the effects of moderate alcohol consumption on foetal death have produced conflicting results and that there appear to be geographical trends in the outcome of such studies. Although a lack of conclusive evidence is a solid basis on which to conduct a new study, the authors do not explicitly detail why the current study will be any different and how it will seek to overcome the limitations of previous research.

The Critical Appraisal Skills Progamme (CASP, www.casp-uk.net) recommends approaching critical appraisals using three steps. The first step is to assess whether the study is valid by evaluating the methodological quality. The methodology of the current study is clearly laid out and replicable. Despite this, one criticism of the methodology is the use of self-reported alcohol consumption data. Self-reported data is vulnerable to social desirability bias by which participants may withhold or fabricate certain behaviours in order to fit in to what others expectations of them. Social desirability has been found to confound reports of other health related behaviours, such as diet (Klesges et al., 2004) and physical activity (Adams et al., 2005). Using the levels of evidence hierarchy (Foster, 2011), cohort studies lay below systematic reviews and randomised control studies in their ability to avoid bias. Therefore, the study’s design helps reduce the risk of other biases that could confound the results. The statistical analysis used was the hazard ratio. This analysis calculates the ratio of the hazard rate corresponding to the two conditions of an explanatory variable (Spruance et al., 2004). In the current study, it was found that women who drank even just low levels of alcohol during pregnancy had higher hazard rates of early foetal death than those women who abstained. This is a suitable statistical analysis that answers the research question at hand.

The second step in the CASP recommendations for appraising evidence is to examine the results. It is important to consider how clinically important the results are and how much uncertainty surround them. Potentially, the current study has excellent clinical utility. Firstly, the sample size was large and as a result, the findings are likely to be very representative of the population as a whole. There is some cultural bias to be aware of as the sample was collected exclusively from a Danish sample. Therefore, the results may not be generalisble to women in other countries. For example, in the UK, alcohol consumption has been found to be much higher, especially among females in their teen years (Mukherjee et al., 2005).

The last step suggested by CASP is to assess whether the results are useful. The current results may certainly be useful in the area of health policy development. The knowledge that even low alcohol consumption within the first trimester of pregnancy can increase the risk of foetal death strengthens current Government guidelines that recommend that women abstain completely from alcohol during pregnancy. In the United Kingdom, the Chief Medical Officer currently advises that women should avoid alcohol altogether but that if they must drink, to consume no more than 1-2 units once or twice a week. However, the current paper suggests that just two drinks per week could increase a woman’s risk of losing the baby. Nevertheless, the conclusions drawn by the study are based on the assumption that alcohol consumption contributed to the increased risk of foetal death and should be interpreted with caution. The study did not collect data on various other variables that may have contributed to the increased risk, such as illegal drug use during pregnancy has been associated with foetal death (Wolfe et al., 2005). Furthermore, consumption of large quantities of caffeine (Wisborg et al., 2003) has been found to be associated with a higher risk of early foetal death. The current study collected information on coffee consumption and it was found that 32.6% of women consumed between one and seven cups of coffee during their pregnancy. Therefore, coffee consumption and not just alcohol may have had an impact on the results. In the discussion, the authors acknowledge the large number of confounding variables that may have impacted upon the study. The discussion of the study does discuss the potential usefulness of the results. However, the authors state in the introduction that discrepancy amongst previous research was a driving force behind the study but fail to discuss why or how the current study may have alleviated this issue.

In conclusion, this is a reasonably strong piece of research that could contribute considerably to health policy. However, it is flawed in some key areas and so the results should be interpreted with caution. For example, if the study is to be replicated, future researchers may consider additional outcome measures that could identify participants at risk of social desirability bias. Nevertheless, the finding that even very low consumption of alcohol before the 16th week of pregnancy may contribute to early foetal death suggests that this should be further investigated as health policy may subsequently consider recommending women abstain completely from alcohol until after their 16th week.

References

Adams, S.A., Matthews, C.E., Ebbeling, C.B., Moore, C.G., Cunningham, J.E., Fulton, J. and Herbert, J.R. (2005) The effect of social desirability and social approval on self reports of physical activity. American Journal of Epidemiology, 161(4), pp. 389-398.

Andersen, A.N., Andersen, P.K., Olsen, J., Gronbaek, M. and Strandberg-Larsen, K. (2012) Moderate alcohol intake during pregnancy and risk of fetal death. International Journal of Epidemiology, 41, pp. 405-413.

Foster, N. (2011) Making sense of the evidential hierarchy. In: Carmen, A. (Ed), Assessing Evidence to Improve Population Health and Wellbeing. Exeter: Learning Matters Ltd.

Klesges, L.M., Baranowski, T., Beech, B., Cullen, K., Murray, D.M., Rochon, J. and Pratt, C. (2004) Social desirability bias in self-reported dietary, physical activity and weight concerns measures in 8-to-10-year-old African-American girls: results from the Girls health Enrichment Multisite Studies (GEMS). Preventative Medicine, 38, pp. 78-87.

Mukherjee, R.A.S., Hollins, S., Abou-Saleh, M.T. and Turk, J. (2005) Low level alcohol consumption and the fetus. British Medical Journal, 330(7488), pp. 375-376.

Spruance, L.S., Reid, J.E., Grace, M. and Samore, M. (2004) Hazard ratio in clinical trials. Antimicrobial Agents and Chemotherapy, 48(8), pp. 2787-2792.

Wisborg, K., Kesmodel, U., Bech, B.H., Hedegaard, M. and Henriksen, T.B. (2003) Maternal consumption of coffee during pregnancy and stillbirth and infant death in first year of life: prospective study. British Medical Journal, 326, pp. 420.

Wolfe, E.L., Davis, T., Guydish, J. and Delucchi, K.L. (2005) Mortality risk associated with perinatal drug and alcohol use in California. Journal of Perinatlogy, 25, pp. 93-100.