The article 'Immunization and Bell's palsy in children: A case-centered analysis', focuses on determining the occurrence of Bell's palsy in children aged 18 or younger and its relationship with immunization. The title of the article is proper as it makes it easier for the reader to understand the problem that the authors are trying to explain. The introductory paragraph of the article outlines the statement of the problem that immunization can be one of the causes for Bell's palsy in children, as more than one case of Bell's palsy has been reported following immunization with influenza and hepatitis B virus (HBV) vaccines. Hence, the authors aim at conducting a population-based epidemiological study of immunization and the relation with Bell's palsy among children aged 18 years or younger. They specifically aim at finding an association between Bell's palsy and HBV vaccine, intramuscular inactivated trivalent influenza vaccines or any other vaccine among children under the age of 18 by using a case-centered analysis. The article begins with an introduction to the condition Bell's palsy, which helps the reader for better understanding of the condition and the incidence of the disease. Overall the authors have done justice to the abstract and introduction section of the article.
The article lacks the literature review part. No specific section is given to the literature review in the article by the authors. They mentioned only about a few studies related to their research following introduction. Hence, the authors neglected to synthesize and review various literatures to make a solid base for their study, but instead jumped directly on the methodology.
Dependent Variable: Bell's palsy, as its occurrence and relationship with immunization is being studied.
Dichotomous dependent variable: Observed immunization status.
Independent Variables: Hepatitis B virus (HBV) vaccine, trivalent influenza vaccine (TIV) and any other type of vaccine, population of Kaiser Permanente Northern California (KPNC), their age, sex, race and socioeconomic status, census distribution and their income. The authors developed inclusion and exclusion criteria for the participants based on the literature.
' Should be diagnosed with Bell's palsy
' Should have one-sided facial paralysis
' The first signs and greatest paresis interval should have been 72 hours
' No history of head injury or cerebrovascular condition, brain tumor, sickle cell disease, Guillain-Barre syndrome or any other neurological signs.
' Absence of any facial muscle weakness or paralysis
' Presence of any other cause of facial paralysis
In this population-based study, the authors have used a case-centered analytical method, which is apparently proper to test the hypothesis developed by the authors. The method of selecting the sample was accurate and properly mentioned in the article. The study sample has been used from the KPNC, where they receive all of their health care at KPNC. All the patient information related to the study was collected through an administrative database system. The authors identified children aged 18 or younger diagnosed with Bell's palsy from first January 2001 to thirty-first December 2006, using the international classification of diseases. Based on the case-centered analysis the authors used a backward approach, where the authors compare the observed outcomes of the immunization during the risk interval that is prior to the onset of an outcome with expected outcomes during the same period of time. In this study authors included samples with only one immunization during the risk interval, which in this case is one year before the occurrence of Bell's palsy. Authors have created specific risk intervals for this study based on the literature and their understanding of the disease, which are 1-2 weeks, 1-4 weeks and 5-8 weeks prior to the onset of the disease. This method, I believe, is appropriate because it minimizes the confusion due to variation in the prevalence of the vaccination and in the occurrence of the BP. Odds ratios were estimated for the relation between BP and immunization. The problem with a case-centered method is that it does not provide proper explanation regarding whether receiving the current vaccination increases the risk of BP or whether it is related to the type and timing of the previous vaccination. Also this study hypothesized that the risk of BP is elevated among vaccinated children who further receive the vaccine within one year. It is not a public health question, rather it should have hypothesized that the risk of BP is elevated among all children following the vaccination. Hence it cannot be generalized to all children. Therefore, although the authors proved their hypothesis, the study lacks the required external validity and is not able to provide accurate conclusion. This method also resulted in loss of 72% of the information. Of the total cases identified, only 28% could be analyzed, because the remaining cases did not get any previous vaccination. Hence, the smaller numbers of cases made the data sparse and thin. Instead of the current performed method, if a nested case-control method would have been performed, most of these limitations could have been prevented. The results would have been statistically more powerful and with no loss of 72% of information, the risks would have been evaluated specifically for TIV and HBV vaccine receivers.
SAS 9.2 was used to conduct the data analysis. Results indicated that total of 977 children were identified with Bell's palsy during the study time period. Depending on the inclusion and exclusion criteria 119 children were rejected as cases and other 36 children were left out due to the previous history of Bell's palsy. Out of the remaining 822 children, based on the inclusion criteria, 233 were included in the study. After performing a descriptive analysis, a logistic regression was performed to find out the significance between immunization with TIV or HBV with Bell's palsy. Results indicated that immunization with either TIV or HBV was not significant with the onset of Bell's palsy during any of the defined time interval of 1-2 weeks, 1-4 weeks or 5-8 weeks. Also no significance was found between any other vaccines with the onset of Bell's palsy. The authors have presented their findings into tables. These tables are well explained and easy to understand for any reader, it is also explained in the text that makes it well organized.
In the discussion section, the authors explain their findings and results of the study. They also provide a superficial relation between their study and other previous studies, but they do not get into any detailed comparison. The authors also provide various strengths and weaknesses of this study. They lack in giving any recommendations for future research.
' Large sample size
' An otolaryngologist to judge al the cases
' Case-centered method lowers privacy concerns in the use of e-records because of lesser number of variables in the final data set
' Small number of cases with intranasal live attenuated influenza vaccine, hence unable to examine the relationship between this vaccine and Bell's palsy
' Lack of standard case definition of Bell's palsy to compare the study with others
' As age and sex has been used as risk factors for this method, other risk factors during the time of immunization cannot be ruled out.
Thus, the study concluded that there is no relationship between the increased risk of Bell's palsy after exposing to the immunization with TIV, HBV or any other vaccines among children aged 18 or younger.
To make the comparison and to prove that the hypothesis of the above mentioned study is correct, I would like to give a brief summary of a similar study.
A study conducted by Cano et.al. (2012), focused on the occurrence of Bell's palsy following the administration of influenza A (H1N1) vaccine. During 2009-2010 influenza seasons Bell's palsy was a major side effect following the 2009-H1N1 vaccines. Results of this study indicated that 65 cases were noted with Bell's palsy following the 2009-H1N1 vaccine and 31 cases were reported after seasonal influenza vaccines. The study concluded that there was no relation between increased risk of Bell's palsy and 2009-H1N1 influenza vaccines.