Community Medicine 2 Output 8 and 9 : PLAN FOR DATA ANALYSIS I. II. Group 15 B, Adviser : Josephine M.

M. Carnate, MD, MPH RESEARCH QUESTION and OBJECTIVES Research Question : Is the prevalence of malnutrition higher in institutionalized children compared to noninstitutionalized children in Paco, Manila and Dasmarias, Cavite during the year 2011? Research Objectives :

General Objective To be able to compare the prevalence of malnutrition among institutionalized children in orphanages and non-institutionalized children located in Paco, Manila and
Dasmarias, Cavite during the year 2011.

Specific Objectives To determine the body mass index (BMI) of institutionalized children in orphanages and non-institutionalized children. To compare the Body Mass Index (BMI) of institutionalized children in orphanages and non-institutionalized children. To correlate the absence or presence of biological parents to the frequency of malnutrition among institutionalized children in orphanages and noninstitutionalized children.

III.

RESEARCH DESIGN

Operational definition of variables:

Dependent variable In this study, the prevalence of malnutrition would be the dependent variable. This variable would be determined using BMI. Values less than 18.5 is regarded as underweight while a BMI value greater than 25 is considered overweight and above 30 is considered obese.as set by WHO.

Independent variables The independent variables in this particular study would be either being an institutionalized child or being a non-institutionalized child. Institutionalized children for this matter will be defined as those living in the orphanages for at least 6 months. Non-institutionalized children are those who are living in a permanent residence with either or both biological parents.

Confounding variables

Confounding variables include age, sex, lifestyle, diet, environment and physical activity.

Research Design:

Type of study to be employed

A cross-sectional type of study will be utilized to identify the relationship between being in an orphanage and/or being in a non-institutionalized child and the prevalence of malnutrition.

Definition of study population/study groups and source of subjects

Children, both male and female, ages 5 to 10 would be the target subjects of this study. They will be randomly selected from a study population which will come from orphanages in Paco, Manila and Dasmarias, Cavite (for the representatives of the Institutionalized children). To represent non-

institutionalized children, the areas within the vicinity of the orphanages will be surveyed to build
the sample population. This is done to eliminate any factors and variables which may affect the outcome of the study like the availability of certain food groups, environmental hazards, degree of pollution et cetera.

Steps to be undertaken/ Procedure to be followed The samples from the aforementioned locations would be randomly numbered and selected using a randomizer (table of random numbers or automated random number generator) to eliminate selection bias. The heights and weights would be determined using a standard balance and height meter respectively. Survey questionnaires would also be given out to the parents and care takers of the orphanages. The data gathered from the questionnaires would be sorted, analyzed and by statistics association of the risk factors involved in acquiring malnutrition for the institutionalized and non-institutionalized children would be done.

Schematic Diagram of the Research Design

Institutionalized Children and Noninstitutionalized Children

Randomly selected children ages 5-10

Institutionalized Child (+) malnutrition Institutionalized Child (-) malnutrition

Biases/Limitations Much consideration had been given to the selection of the ages from 5 to 10 primarily which is the facilitation of the measuring of the heights and weights to determine accurate Body-mass indices. This range has also the advantage of being able to confirm and validate certain information from the survey questionnaires which in most cases will be answered by the parents or care takers of the orphanage. Likewise, the degree of cooperation tends to increase from ages 5 and up. Therefore the study was decided to be limited to this age range. Another limitation to this study is the inability of the researchers to determine and eliminate confounders like inherent defects in metabolism which may significantly affect the outcome of the study. Likewise, the difficulty of defining and determining factors such as crowding limits the study to some extent. Other quantitative determinants of malnutrition will also not be included in this study which will solely rely on BMI data to report malnutrition. The extent of association between the availability of some food groups, presence of pollution and other variables would not be discussed intensively but their individual possible contribution to malnutrition would be thoroughly discussed. Selection bias may follow the tendency of the researchers to opt for malnourish-looking individuals in hopes of generating a large sample population with a very high percentage of positive malnutrition. This will be circumvented by using random selection of samples from the representative population.

IV.

PLAN FOR DATA ANALYSIS

Descriptive Statistics Measures of Disease frequency

The most important variable that will be measured in this study is the BMI of all sets of subjects. To be able to determine the frequency of malnutrition among the subjects, one can present the BMI values in the ordinal scale (underweight, normal, overweight, obese) or using the ratio scale of measurement. However, in order to measure the discrepancies between the values and analyze further the reasons and the multiple probable factors behind the computed differences, this study will make use of the ratio scale. This will also allow for a lot more values for the frequencies of distribution which again will be helpful in the analysis and interpretation of all the data gathered. Since the subjects are also randomly selected, nominal variable GENDER and the ratio value AGE may provide interesting correlations and thus their respective frequencies of distribution would be gathered as well. The number of subjects for the exposed (institutionalized) and the unexposed (non-institutionalized) population have equal frequencies of distribution. If available, and for the purpose of analysis and discussion, the number of months of stay in the institution will be collected as well and their relative frequencies will be cross-matched with the degree of malnutrition, if there would be. Measures of Central Tendency

Measures of central tendencies are statistical values which best represents an entire group of values or distribution. Mean is defined as the sum of all values divided by the total number of observations. Median is the midpoint in a set of ordered values or measurement. It is the point at which 50% of the values fall above and 50% fall below. It is obtained by arranging the values from smallest to largest and finding the middle value. If the number of observations is odd, the middle

value is the median. If the number of observations is even, the median is the average of the two middle values. Mode, the third measure of central tendency, is the category or value that occurs most frequently in a distribution. For the study at hand, in analyzing the values obtained in the determination of BMI, the above measures of central tendency may not be effective descriptors. This will be illustrated by the following examples. Suppose 5 BMI values were obtained by calculating the height and weight measurements of 5 subjects and the values obtained were 17, 17, 22, 27, and 27. If the arithmetic mean would be calculated, the value that the researcher would arrive at would be 22. The BMI range used in this particular study set by WHO as normal would be 18.5 25. If the researcher would use the mean to describe the set of data, he would be concluding that generally, the subject population falls within normal BMI range. This is obviously not the case. The same thing would be true if the median would be utilized. Now suppose another set of BMI values was obtained: 17, 17, 17, 19, 20, 20, and 21. If the mode would be used to describe the set, the researcher would arrive at a conclusion which states that generally, the subjects in this particular set are below normal. Again, this is simply not the case and therefore the use of mode is not warranted. For the variable age, the arithmetic mean will be of help in giving the representative age of the population.

Measures of Dispersion

Statistical dispersion is variability or spread in a variable or a probability distribution. Common examples of measures of statistical dispersion are the range, variance and standard deviation which are all measures of absolute dispersion, and coefficient variation which is a measure of relative dispersion. Range is the length of the smallest interval which contains all the data. It is calculated by subtracting the smallest observation (sample minimum) from the greatest (sample maximum). Variance describes how far the numbers lie from the mean (expected value). It forms part of a systematic approach to distinguishing between probability distributions. Standard Deviation shows how much variation or dispersion exists from the average (mean, or expected value). A low standard deviation indicates that the data points tend to be very close to the mean, whereas high standard deviation indicates that the data points are spread out over a large range of values. Coefficient of variation (CV) is a normalized measure of dispersion of a probability distribution. It is also known as unitized risk or the variation coefficient. The absolute value of the CV is sometimes known as relative standard deviation (RSD), which is expressed as a %. It is useful in comparing the results obtained by different persons who are conducting measurements involving the same variable but different scales of measurement (QOL using two different indices). For this particular study, as is the case with most medical studies, Standard Deviation would be employed. This has the advantage of measuring the average deviation of the values from the mean BMI of a particular set. This would be beneficial in the interpretation of data specially when multiple factors will be considered.

Measures of Association

An association is present if probability of occurrence of a variable depends upon one or more variable. There are different measures of association used for different specific studies such as Odds Ratio, Relative Risk Ratio and Prevalence Rate Ratio.

1. Odds ratio is a measure of effect size, describing the strength of association or nonindependence between two binary data values. It is used as a descriptive statistic that treats the two variables being compared symmetrically, and can be estimated using some types of non-random samples. The odds ratio is commonly used for case-control studies, as odds, but not probabilities, are usually estimated. Odds ratio is directly proportional to the odds of exposure among those with the disease (cases) and indirectly proportional to the odds of exposure among those without the disease (controls). 2. Relative risk is a ratio of the probability of the event occurring in the exposed group versus a nonexposed group. Relative risk is used frequently in the statistical analysis of binary outcomes where the outcome of interest has relatively low probability. It is thus often suited to clinical trial data, where it is used to compare the risk of developing a disease. Relative risk is used in randomized controlled trials and cohort studies. It is directly proportional to the incidence of disease among the exposed and indirectly proportional to the incidence of disease among the unexposed. 3. Prevalence Rate Ratio is used as an estimate of how common a disease is within a population

over a certain period of time. It is directly proportional to the prevalence of the disease among the exposed and indirectly proportional to the prevalence of the disease among the unexposed.

Our research study entitled Comparison of Malnutrition among Institutionalized and NonInstitutionalized Children in Paco, Manila and Dasmarinas, Cavite as determined by Body Mass

Index , will use cross-sectional research design thus Relative Risk Ratio as measure of association will be utilized. This can be done by tabulating the data gathered from the actual data collection.

Malnutrition + + Exposure Total C A+C D B+D A B -

Total

A+B C+D A+B+C+D

Relative Risk =

A A+B C C+D

After computation, the result will be interpreted based on the following: RR = 1 there is no association RR 1 there is an association RR > 1 (+) causative association RR < 1 (+) preventive association

Inferential Statistics

Plan of hypothesis test applicable to the data set A completely randomized t-test is used to determine whether or not two independent populations have different mean values on some measure. This type of statistical data analysis procedure to hypothesis testing will be used to be able to determine the differences in the prevalence of malnutrition in institutionalized and non-institutionalized children. The dependent variable, which is the incidence of malnutrition among the institutionalized and non-institutionalized children, will be accounted for upon the computation of the Body Mass Index of the sample population for both the institutionalized and non-institutionalized children. These values will then be averaged to get the mean value of malnourished children among the institutionalized and non-institutionalized children to be able to compute for the t-value. The t-value will determine if the ratio is large enough to say that the difference between the groups is not likely to have been a chance finding.

Possible interpretation of data

Much of the discussion in the possible interpretation of data had been covered in the length of this paper. Salient points need reiteration however. For the determination of the frequency of malnutrition among the subjects, the BMI would be tabulated and the respective frequencies of distribution would be determined after which a comprehensive classification of the BMI values would be used to sort the data out. The classified values would be important not just in setting what is normal, but also in determining the degree of its deviation from the normal, that is, the severity of malnutrition.

The research hypothesis can also now be tested by determining the RELATIVE RISK RATIO using the formula provided for in the above section titled Measures of Association. The confidence interval would be calculated and set and if a value of more than one is obtained (and assuming a value of < 1.0 for the lower limit of the confidence interval) then there would be sufficient evidence to conclude a causative association between being in an orphanage and malnutrition. The same process would be used in calculating for the association between being non-institutionalized and malnutrition. In the event that both sets (institutionalized and non-institutionalized) give the same association (either protective or causative) then factors like food, lifestyle, economic status and others would be analyzed with the help of survey question forms provided. Also as stated above, age and gender would also be cross-matched with the occurrence of malnutrition using Epi Info 7.0 to take into consideration their possible contribution to having or not having malnutrition.