Critical Appraisal of a Research Paper on the Association between Obstetric Complications and Risk of Schizophrenia

This critical appraisal evaluates a research paper on the association between obstetric complications and the risk of schizophrenia. It examines the study design, findings, internal validity, and external validity. The paper concludes that while the study provides some insights, there are limitations that affect its validity and generalizability.

Running head: EPIDEMIOLOGY
Epidemiology
Name of the student:
Name of the University:
Author’s note

1NURSING
Introduction:
Critical appraisal is the systematic process of review and analysis of research findings
based on research question and assessment of trustworthiness, value and relevance of research
for a particular context. Critical appraisal is an important skill to engage in evidence based
practice and find ideas for improvement in quality of care. It emphasized on the review of a
research paper to identify whether it can facilitate practice improvement and application in target
setting or not (Elwood, 2017). The main purpose of this report to conduct a critical appraisal of
the research by Byrne et al. (2007) and evaluate the research paper based on internal and external
validity and analyzes causal association between exposure and outcome.
Overview of the paper:
The main purpose of the research article by Byrne et al. (2007) was to investigate about
the risk of schizophrenia due to exposure to range of obstetric events. The paper is focusing on
the issue of obstetric complications (OC) like abnormal fetal growth, pregnancy complication
and complications of delivery. The paper justifies the need for conducting such study by giving
evidence regarding the risk of schizophrenia due to OC. The literature review suggested
prevalence of OCs during pregnancies and interaction of OCs with family history to increase risk
of schizophrenia. It also suggested lack of clarity regarding impact of OCs on schizophrenia risk.
The study was done in 2007 and the literatures between 1999 to 2002 were used to indicate gap
in past research and justify the rationale for doing the research. Current research literature
suggests that environmental and genetics factors play a role in etiology of schizophrenia. The
risk of schizophrenia increases due to perinatal factors like OC, environment and parental
influence on the fetus during pregnancy (Jenkins, 2013). The novel approach that the study took
was that to confirm the risk of the disease due to OCs, the study aimed to explore confound
effects of socioeconomic factors and family history of psychiatric illness. The study helped to
define whether parental aged, socioeconomic status, family history, place of birth and maternal
citizenship leads to development of schizophrenia or not.
Exposure or intervention for the study:

2NURSING
Although the paper does not evaluate any intervention, however the main exposure that
the study is evaluating includes risk of schizophrenia due to exposure to OCS. The study
compared different demographic factors to predict risk of schizophrenia.
Study design:
The study used national population nested case-control study to investigate about the
impact of range of OCs in risk of schizophrenia. Nested case-control research design includes
that research design in which case and control groups are taken from a recognized cohort. The
case groups involved those sample groups which are exposed to exposure of interest and control
group consist of controlled case. The advantage of using this research design is that it
particularly measures time dependent variables. However, the disadvantages of nested case-
control study includes reduces precision and chances of flaws in the sample design (Breslow &
Weiss, 2018). Since the paper aims to evaluate exposure to OC, use of this design is appropriate.
The Danish national register was used to collect information on case population and control
population
Study population:
The study was based on two population group which is the case and control population.
The case population consisted of group who were exposed to OC and schizophrenia. From the
review of persons above 15 years visiting Danish psychiatric facility between 1981 and 1998,
total of 1039 persons with schizophrenia were identified. In addition, the criteria for selection of
control population include those who never visited psychiatric facility.
Main findings and future directions:
The review of result finding revealed no statistically significant association between
obstetric factors and family history of psychiatric illness was found. In addition, the comparison
of OC with other confounding factors revealed very moderate associated maternal behavior and
maternal infections on risk of later development of schizophrenia. The review of national
population data revealed that risk significantly increased when mothers revealed no previous
antenatal visits, experiences of premature deliver and sepsis of child birth. The paper did not
directly discussed about the key implications of the research findings. However, it suggested the

3NURSING
need for future research by recommending the need to develop better understanding regarding
role of obstetric events in risk of schizophrenia. The link with development trajectory of women
is a major finding of the research.
Internal validity of the paper
The internal validity of a research paper is the process used to enhance the quality of
research by focusing on addressing biases, confounding factors and other internal factors that
could have an impact on research design. Internal validity is reviewed by analyzing the research
paper and identify whether the researcher implemented steps to control the extraneous variables
that influences study outcome or not (Patten & Newhart, 2017). Hence, in case of studies where
causal association between exposure and an outcome is being investigated, consideration of
factors that could influence outcome of interest is important. For the paper by Byrne et al.
(2007), the internal validity of the study has been considered by critical appraisal on clear
presentation of research findings, possibility of chance variation, possibility of bias and impact
of confounding factors on research.
As the study mainly explored the impact of demographic factors on risk of schizophrenia,
the research finding was grouped into those born in the year 1973-1983 and 1978-1983. By
representation of data related to prematurity, manual extraction, length at birth and birth weight,
it was found that risk for schizophrenia is associated with fetal growth and maturity factors. The
presentation of data for each age group revealed that different types of OCs strongly predicted
risk of schizophrenia. Three or more previous pregnancies compared to other single pregnancy
were the highest risk factor of schizophrenia. However, no relation was found between
demographics factors like parental age, place of birth and maternal citizenship does not influence
outcome. Hence, the research clearly answers the research question.
Despite the above results, the validity of the study finding is also influenced by presence
of many confounding factors and biases in research design. Firstly, the use of national register to
evaluate causal association contributes to biases because it did not followed single clinical
guideline to diagnose risk of schizophrenia. The register had diagnoses based on either
ICD(International Qualification of Disease)-8 or ICD-10. Hence, this contributes to biases
because to promote uniformity in research data, it was necessary to engage in take cases from the

4NURSING
time when diagnoses via ICD-10 started. To enhance the internal validity of data and reduce
impact of bias on research outcome, it was necessary to used update diagnostic tool. This would
have helped to reduce impact of confounding factor on research outcomes.
The review of sample recruitment data also indicated about presence of confounding
factors and chance variation in research. This can be said because the eligibility criteria for
sample recruitment were recruitment of people above 15 years. The research findings was also
distributed according to findings obtained for two different date of birth for samples. The main
rationale behind this was to evaluate the impact of maternal age at birth on risk of schizophrenia.
However, limitation or biases in this approach is that the sample mainly includes samples with
young people mainly 19 years for the study. Because of this form of biases, the study finding
cannot be generalized to all people with schizophrenia. According to Coid et al. (2017), time and
place of birth of birth is a predictor of greater risk of schizophrenia. Moving from urban to rural
area increase risk of schizophrenia. In contrast, shifting to rural location reduces the likelihood of
developing the mental disorder. However, the paper regarded urbanicity as negligible contributor
to identify risk ratio in case group. Hence, ignoring this factor contributes to change variation
and lack of reliability of research findings. Sub-group analysis was conducted and interaction
between relevant research variables like OCs and sex and family history of mental illness was
identified. However, gap found in evaluating interaction was that family psychiatric history was
evaluated based on cases who are admitted to hospital. This is not reliable data as it represents
only very serious cases and does not give overview about actual rate of psychiatric history in
sample groups. Hence, these biases affect the reliability of the study.
Causal association between exposure and outcome
The main aim of the study was to explore association between exposure (OCs) and
outcome (schizophrenia). In the field of epidemiology and in epidemiological trials, causal
association is determined by association between risk factor and outcome. This form of
association is mostly confirmed by observation and experimental research design. However,
observation of association may not be true because of the presence of chance errors, biases and
confounding factor (Vandenbroucke, Broadbent & Pearce, 2016). However, one of the major gap
found in the paper included lack of use of observation or experimental method to investigate
about the research topic. Taking data from national register is one of the major limitations.

5NURSING
A valid causal association is one in which authors considers all risk factors that could
contribute to health outcomes. In the context of risk factor of schizophrenia, the study evaluated
the effect of many risk factors. This included considering effect of parental socioeconomic
status, age of parents, urban place of birth and family history of psychiatric disorder. Age of
parents was not found to be a risk factor for schizophrenia. Instead of parental age, the rate of the
disorder was influenced by frequency of previous pregnancy and experiences of OCs. The
strength of the strategies used for exploring causal association is that it used the right risk factor
to investigate about the research topic. A study by Wu et al. (2018) investigating about the
epidemiology and risk factors schizophrenia revealed that clinical, behavioral and demographic
information are vital source to determine prevalence and incidence of schizophrenia. For
example aggressive behaviors have been highly regarded as a social problem that helps to predict
diagnoses of schizophrenia as well as severity of the disease. However, for the paper by Byrne et
al. (2007), lack of attention to other possible risk factors such as behavioral data and clinical
symptoms was a factor that acted as a barrier in getting a plausible causal association. In
addition, presence of confounding factors like collecting only admitted cases for family
psychiatric history further affected the validity of the demographic risk factors comparing causal
association.
The confounding factors could have been minimized if the author had taken the steps to
identify causal association based on several criteria or evidence based criteria. The author
identified OCs as the major risk factor of OCs. However, OC is a marker of schizophrenia.
However, it is not the causal factor in the etiological pathway for schizophrenia (Forsyth et al.,
2012). Therefore, focusing on gene-environment interactions and maternal factors would have
been a true consideration of risk factors. The quality of the study is affected by lack clear
consideration of causal association between OCs and schizophrenia.
External validity of the paper:
External validity of a research involved applying all possible steps to develop valid
conclusion regarding the research finding and evaluate the extent to which the study can be
generalized across target setting and audience. External validity can also be defined as the extent
to which research findings can be generalized to a large group. Some of the factors that can have
an impact on the external validity of the study includes population characteristics, data collection

6NURSING
method, effect of research environment and relation between study subject and research
participants.
The research paper collected data from a Danish population register. The main criteria for
sample selection was to take only individual aged above 15 years and those with schizophrenia
were selected to identify whether diagnosis of the disease is influenced by demographic factors
or not. The characteristics of study population suggest that the average sample year for the group
included 19 years and it represents only those patients who were admitted for a psychiatric
problem. This form of sample characteristics may affect the external validity of the study. This is
said because this sample group is not representative of entire population group who have been
diagnosed with schizophrenia. By taking only young age group from national register instead of
evaluating findings from experimental research, the study missed many other risk groups who
are at risk of schizophrenia. Hence, because of this limitation, the study outcome cannot be
generalized for all people with schizophrenia. Khorsan and Crawford (2014) argues that patient’s
baseline characteristics like age, gender and severity of diseases geographical setting and health
care systems must be uniform to minimize external barriers to the application of research
evidence.
The external validity of the study is also hampered by lack of uniformity in using tools to
identify patients with risk of schizophrenia. The study findings are most likely to disseminated to
target setting so that appropriate changes are implemented in promote quality of care and
practice improvement. However, the likelihood of dissemination of findings of this study is
limited because of poor consideration to take a valid diagnostic tool to confirm incidence of
schizophrenia in case group. The review of the article revealed that data was taken for those
entries which were done either by ICD8 or ICD 10. Taking both of the guideline for recruiting
patient is a major limitation as there are differences in diagnostic criteria for ICD 8 and ICD 10.
Hence, the population under study in the research group may not be representative of actual
patient at risk of schizophrenia. This is said because the process used for diagnosing risk is not
congruent with current practice related to diagnosis of the disorder. Hence, these factors may
seriously affect the purpose of investigation and limit application of findings in actual setting.
The sample size was adequate, however as it was not representative of target population of
interest, the external validity of the study is low.

7NURSING
Conclusion:
From the critical appraisal of the research paper on exploration of association between
OCs and risk of schizophrenia using nested case control study, it can be concluded that the study
gave answer to the research question. The study revealed no interaction between sex and
psychiatric history. However, OCs and maternal infection and hypoxia were a moderate predictor
of risk. Although the research finding demonstrates causal association between exposure and
outcome, however the validity and reliability of the study is affected by presence of many
limitations. Firstly, the study is found to be contributing to biases and confounding factors by
taking only data from Danish longitudinal registers for fulfilling the aim of research. Taking this
source is a major reason behind high risk of confounding factors and chances errors in the
findings. In addition, other factor that limits the quality of study includes lack of considerations
to detect psychiatric history of patients. By taking data from admitted cases list, the study
findings may be reflective of only severe cases of schizophrenia. Such biases in data collection
affects the external validity of the study findings too.

8NURSING
References:
Breslow, N. E., & Weiss, N. (2018). Nested Case-Control Studies: A Counting Process
Approach. In Handbook of Statistical Methods for Case-Control Studies (pp. 329-349).
Chapman and Hall/CRC.
Byrne, M., Agerbo, E., Bennedsen, B., Eaton, W. W., & Mortensen, P. B. (2007). Obstetric
conditions and risk of first admission with schizophrenia: a Danish national register based
study. Schizophrenia research, 97(1-3), 51-59.
Coid, J. W., Hu, J., Kallis, C., Ping, Y., Zhang, J., Hu, Y., ... & Kirkbride, J. B. (2017). Urban
birth, urban living, and work migrancy: differential effects on psychotic experiences
among young Chinese men. Schizophrenia bulletin, 44(5), 1123-1132.
Elwood, M. (2017). Critical appraisal of epidemiological studies and clinical trials. Oxford
University Press.
Forsyth, J. K., Ellman, L. M., Tanskanen, A., Mustonen, U., Huttunen, M. O., Suvisaari, J., &
Cannon, T. D. (2012). Genetic risk for schizophrenia, obstetric complications, and
adolescent school outcome: evidence for gene-environment interaction. Schizophrenia
bulletin, 39(5), 1067-1076.
Jenkins, T. A. (2013). Perinatal complications and schizophrenia: involvement of the immune
system. Frontiers in neuroscience, 7, 110.
Khorsan, R., & Crawford, C. (2014). External validity and model validity: a conceptual approach
for systematic review methodology. Evidence-Based Complementary and Alternative
Medicine, 2014.
Patten, M. L., & Newhart, M. (2017). Understanding research methods: An overview of the
essentials. Routledge.
Vandenbroucke, J. P., Broadbent, A., & Pearce, N. (2016). Causality and causal inference in
epidemiology: the need for a pluralistic approach. International journal of
epidemiology, 45(6), 1776-1786.

9NURSING
Wu, Y., Kang, R., Yan, Y., Gao, K., Li, Z., Jiang, J., ... & Xia, L. (2018). Epidemiology of
schizophrenia and risk factors of schizophrenia-associated aggression from 2011 to
2015. Journal of International Medical Research, 46(10), 4039-4049.