NYU Week 8 Precede Proceed Model & Confounding Variables and Bias Discussion Responses
The PRECEDE-PROCEED model (PPM) is a process that focuses on the community as a source of public health promotion. This model is used as a comprehensive road map of directions for constructing, applying, and assessing health promotion and health intervention methods of health. Proceed stands for policy, regulatory, and organizational constructs in educational and environmental development, which is the structure for applying and assessing public health plans. Proceed focuses on 3 phases, such as process evaluation, impact evaluation, and outcome evaluation. Process evaluation determines if the intervention methods are reaching the targeted community and are accomplishing the goals that were set. Impact evaluation evaluates if there is behavioral change and outcome evaluation identifies if the intervention is reducing the prevalence of the negative behavior and increasing the positive behavior (Glanz et al., 2015; Rural Health Information Hub, 2020).
Precede stands for predisposing, reinforcing, and enabling constructs in educational diagnoses and evaluations, which is the structure for planning a specific public health plan. Precede focuses on 5 phases such as (1) social assessment, (2) epidemiology/behavioral and environmental assessment, (3) educational/ecological assessment, (4) administrative and policy assessment and intervention alignment, and (5) implementation. Phase 1 determines the social problems and needs of a community and identifies the results. Phase 2 identifies the health determinant issues when the problem is identified and set goals towards it. Phase 3 analyzes the behavioral and environmental determinants that enhance the identified problem. Phase 4 influences the implementation of interventions that encourages positive behavior change. Phase 5 focuses on implementing interventions (Glanz et al., 2015; Rural Health Information Hub, 2020).
The intervention mapping model (IMM) builds on the logic model of the problems that were identified, developed and expanded on in the PRECEDE-PROCEED model. This model works as a guide that focuses on interpreting the determinants that affect behavioral and environmental change and matches theory-based changed methods to the determinant to improve and promote health. This model has 6 steps, such as (1) logic model (theory) of the problem, (2) program outcomes and objectives-logic model (theory) of change, (3) program plan, (4) program production, (5) implementation plan, and (6) evaluation plan (Glanz et al., 2015).
Step 1 uses a planning group to describe the framework of the intervention and conduct a needs assessment to create a logic model of the problem. Step 2 uses theory and evidence to define and identify the targets of change for the intervention. Step 3 focused on generating a program theme that matches a theory-based and evidence-based intervention method and then selecting the practical application that will deliver the intervention change. Step 4 incorporates methods and practical applications into a structured program by developing the needed materials and messages. Step 5 is the planning phase that focuses on the implementation, adoption, maintenance, and sustainability of designing and testing the effectiveness of the intervention program. Step 6 creates an evaluation plan that conducts the outcome and process evaluation method (Glanz et al., 2015).
The PRECEDE-PROCEED model is the foundation for the intervention mapping model. Both models are used as community-based guides that use the ecological approach in planning and providing guidance for the use of theories to understand the health problems and plan the intervention methods that are caused by determinants factors. Additionally, I prefer the PPM because its steps are more detailed, easier to use on other health problems, it provides flexibility and is appropriate to use on a larger scale for various health problems (Glanz et al., 2015).
An example of the PPM is the child pedestrian injury prevention project. The targeted group is children age 5-10 and the epidemiological factor is severe head injuries for child pedestrian injuries. The behavioral risk factors are children lacking parental supervision when crossing, and children lacking education on road crossing procedures. The environmental risk factors are roadside obstacles such as road construction, and high traffic volume and speed, such as rush hour traffic. The contributing factors are parents allow their children to cross the roads alone. Intervention methods are education on road crossing procedures at school and at home implementing safety traffic signs near school zones.
An example of the IMM is using the “It’s your game” program (sexual health education program) to reduce adolescent risky sexual behavior that results in STI/STD. The targeted group is predominantly African American and Hispanic middle school students in urban school districts with a determinant factor of low socioeconomic status. The behavioral factor is risky sexual behavior for STI/STD and teen pregnancy and environmental factor is lack of parental communication about sexual health and lack of parental monitoring. The next step is to design a plan related to the determinant that would influence change in persuading the children not to have sex, use a condom, use contraceptives, and get tested. The plan is to use computer activities and interactive programs with feedbacks to reinforce and enable students from partaking in risky behavior. Next is implementation by working with the community to provide free condoms in school and teaching teachers how to deliver the “it’s your game program.” Lastly, to evaluate the programs’ progress, conduct an experiment that measures the intervention outcome such as testing the students’ knowledge on STI/STD (Glanz et al., 2015).
Reference
Glanz, K., Rimer, B. K., & Viswanath, K. (2015). Health behavior: Theory, research, and practice (5th ed.). San Francisco, CA: Jossey-Bass.
Respond to the bold paragraph ABOVE by using one of the option below…
- Ask a probing question.
- Share an insight from having read your colleague’s posting.
- Offer and support an opinion.
- Validate an idea with your own experience.
- Make a suggestion.
- Expand on your colleague’s posting.
QUESTION 1
Bias is a form of prejudice where a group or individual is treated unfairly compared to others. For example, racial bias, which is one race being treated unfairly in a situation compared to another race. Confounding is when distortion is created between the exposed and outcome because of another variable. The criteria for establishing a confounding is the connection between if exposure A is a cause of the disease B. For example, is the factor a known risk for causing the disease B and is the factor associated with exposure A, but is not a result of exposure A. For example, during a study, it showed that increased coffee drinking increases the risk of pancreatic cancer. Pancreatic cancer is diagnosed with 28,000-30,300 individuals per year with a low survival rate of 19% in 1-year and 4% in 5 years. However, due to confounding, a third variable appeared, which is smoking, such as increased coffee drinking leads to smoking, which leads to increased risk of pancreatic cancer. This shows that the confounding factor affects the diseases but can be associated or not with coffee drinking (Celentano& Szklo, 2019; center for diseases control and prevention, 2019).
A way to establish adjustment in a confounding relationship is stratification when the number of factors available is limited. For example, the textbook demonstrated a situation where smoking is a risk for lung cancer, but air pollution could be a factor. Lung cancer is a serious illness that has many factors, where approximately 235,000 new cases of lung cancer appears each year, with 14% of newly discovered cancers being lung cancer, with approximately 155,000 deaths per year. Therefore, using stratification, the researchers compared the rate of lung cancer in urban community in smokers and nonsmokers. If the rate of lung cancer is higher in smokers compared to nonsmokers, then smoking is the factor. However, if more lung cancer rate is higher in nonsmokers than smokers, then air-pollution is the factor (Celentano& Szklo, 2019; American cancer society, 2020).
Reference
American Cancer Society. (2020). Key statistics for lung cancer. Retrieved from https://www.cancer.org/cancer/lung-cancer/about/key-statistics.html
Celentano, D. D., & Szklo, M. (2019). Gordis epidemiology (6th ed.). Philadelphia, PA: Elsevier. ISBN-13: 9780323552295
center for disease control and prevention. (2019). Pancreatic Cancer. Retrieved from https://www.cdc.gov/niosh/nioshtic-2/20037430.html
QUESTION 2
Bias and confounding in epidemiological study may reflect the true effect of an exposure on the development of and outcome under investigation, this should always be considered that the findings may be the fact that is due to an alternative problem. If the alternative may be due to the effects of a random error, bias or confounding may produce spurious results, leading us to conclude the existence of a valid statistical association when one does not exist or alternatively the absence of an association when one is truly present (Carneiro I, et al, 2011). While in the observational studies are particularly susceptible to the effects of chance, bias and confounding and these factors need to be considered at both the design and analysis stage of an epidemiological study so that their effects can be minimized. Bias is a systematic error in an epidemiological study that results in an incorrect estimate of the true effect of an exposure on the outcome of interest. It can also result from systematic errors in the research methodology, of which the estimate is either above or below the true value, depending on the direction of the systematic error (Choudhary, P., et al, 2020). Next is the magnitude of bias is generally difficult to quantify, and limited scope exists for the adjustment of most forms of bias to analysis stage. As a result, careful consideration and control of the ways in which bias may be introduced during the design and conduct of the study is essential in order to limit the effects on the validity of the study results. There are different types of bias that have been identify in epidemiological studies, but for simplicity they can be broadly grouped into two categories: information bias and selection bias. While confounding is Confounding provides an alternative result for an association between an exposure and an outcome. It occurs when an observed association is in fact distorted because the exposure is also correlated with another risk factors. This risk factor is also associated with the outcome, but independently of the exposure will be under investigation, as a result, if the consequence, the estimated association is not that same as the true effect of exposure on the outcome. An unequal distribution of the additional risk factor, between the study groups will result in confounding. The observed association may be due totally, or in part, to the effects of differences between the study groups rather than the exposure under investigation. A potential confounder is any factor that might influence the risk of disease under study (Dechao Feng, et al, 2020). This may include factors with a direct causal link to the disease, as well as factors that are measures for other unknown causes, such as age and socioeconomic status.
For a variable to be considered as a confounder, it must be associated with the outcome, example, a person who is at risk. Next the variable must also be associated with the exposure under study in the source population and should not be lie on the causal pathway between exposure and disease outcome. When studying alcohol use consumption to be associated with the risk of coronary heart disease (CHD). However, smoking may have confounded the association between alcohol and CH. Smoking is a risk factor for CHD, so is independently associated with the outcome, and smoking is also associated with alcohol consumption because smokers tend to drink more than non-smokers. Controlling for the potential confounding effect of smoking may in fact show no association between alcohol consumption and confounding factor, if not controlled for, cause bias in the estimate of the imp when act of an exposure. Effects of confounding may result in observed association when no real association exists, no observed association when a true association does exist and when an underestimate and overestimate of the association is in a negative and positive confounding (Hennekens CH, et al, 1987). Confounding can be addressed either at the study design stage or adjusted to analysis stage to provide enough relevant data that have been collected. Several methods can be applied to control the potential confounding factors and aim that will make the groups as similar as possible with respect to confounding cases. Potential confounding factors may be identified at the design stage that is based on previous studies or link between the factor and outcome may be considered as biologically Methods to limit confounding at the design stage include randomization, restriction and matching ( Choudhary, P.,et al, 2020).
This is the ideal method of controlling for confounding because all potential confounding variables, both known and unknown, should be equally distributed between the study groups. It involves the random allocation (e.g. using a table of random numbers) of individuals to study groups. However, this method can only be used in experimental clinical trials. In restriction it helps limits participate in the study of an individuals who relate to the confounder. For example, if participate in a study is restricted to non-smokers only, any potential confounding effect of smoking will be eliminated. However, a disadvantage of restriction is that it may be difficult to generalize the results of the study to the wider population if the study group is homogenous. While matching involves selecting controls so that the distribution of potential confounders, such as age or status of those that smoke is as similar as possible to other case. In practice this is only utilized in case-control studies, but it can be done in two ways, like pair matching, selecting for each case one or more controls with similar characteristics of the age of smoking habits. Also, the frequency matching is to ensure that as a group the cases have similar characteristics to the controls (Xiao Hu, et al, 2020).
References:
Choudhary, P., & Nain, N. (2020). CALAM: model-based compilation and linguistic statistical analysis of Urdu corpus. Sadhana, 45(1), 1–10. https://doi-org.lopes.idm.oclc.org/10.1007/s12046-019-1237-3
Dechao Feng, Xiao Hu, Yin Tang, Ping Han, & Xin Wei. (2020). The efficacy and saety of miniaturized percutaneous nephrolithotomy versus standard percutaneous nephrolithotomy: A systematic review and meta-analysis of randomized controlled trials. Investigative & Clinical Urology, 61(2), 115–126. https://doi-org.lopes.idm.oclc.org/10.4111/icu.2020.61.2.115
Hennekens CH, Buring JE. Epidemiology in Medicine, Lippincott Williams & Wilkins, 1987.
Carneiro I, Howard N. Introduction to Epidemiology. Open University Press, 2011.
QUESTION 3
There are various forms of bias, but the two major types are selection bias and information bias. Selection bias is when a group or individual is selected, and it is apparent that their selection is not due to randomization. For example, during a case study, 20 individuals were asked to participate in a survey, where 8 individuals are positive for a disease and the other 12 are negative. In this case, the research would use selection bias to pick the 8 positive individuals to prove their data, but it is obvious that randomization did not work. Information bias is when inaccurate data is measured to give flawed data. For example, during a controlled experiment, some of the controlled individuals were classified as exposed, while some of the exposed individuals were classified as the control, which makes the data for the experiment inaccurate (Celentano& Szklo, 2019).
The article I chose is called how to avoid bias in systematic reviews of observational studies by Almeida & Goulart. The article discussed how bias emerges in a systematic review by selection bias, information bias, and confounding bias. To minimize these occurrences of bias, the funnel plot technique (visual aid in detecting bias), statistical tests, and using communication tools to get in contact with researchers to get more details on primary data is effective. Utilizing more than one reviewer to independently collect and analyze data from primary studies and using comparable methodological quality to allow the identification of biases (Almeida & Goulart, 2017).
Additionally, bias can originate due to multiple steps of an investigation or from a retrospective cohort where bias arises due to missing information due to using existing records or selecting the individuals after an outcome has already occurred. To prevent this researchers must try their best in improving the design and applying statistical techniques when evaluating the results. Establishing a follow-up procedure to keep records of all variables to prevent missed information. Designing studies by including confounding variables from the beginning, matching by age and gender and including adjustment techniques (Ramirez-Santana. 2018).
Inferences based on data with bias provides inaccurate results with missing information. Therefore, the result will not be valid or credible for use. Additionally, if the data interpreted was biased with false information and that biased result was used to make a drug, people ingesting that drug can be killed, because the formula might be incorrect and the drug might be deadly for the human body or the dosage might be too strong/lethal.
Reference
Almeida, B.P.C., & Goulart, G.N.B. (2017). how to avoid bias in systematic reviews of observational studies. Revista Cefa, 19(4), 551-555. Doi: https://doi.org/10.1590/1982-021620171941117
Celentano, D. D., & Szklo, M. (2019). Gordis epidemiology (6th ed.). Philadelphia, PA: Elsevier. ISBN-13: 9780323552295
Ramirez-Santana, M. (2018). Limitations and Biases in Cohort Studies. Retrieved from https://www.intechopen.com/books/cohort-studies-in-health-sciences/limitations-and-biases-in-cohort-studies
