Capstone Projects - Asia
Name: Sarah Hurd MS '12
Degree and Year: Kinesiology and Pre-Physical Therapy
Kansas State University, 2009
Capstone Project: Determining the efficiency of government primary health care centers in Southern Karnataka, India
Abstract: Data envelopment analysis (DEA) is a useful, practical method for evaluating the efficiency of decision-making units (DMUs) in comparison to the best practices of their peers. This method has been used extensively in healthcare around the world to evaluate the effectiveness of resource allocation and to ensure the conservation of limited resources. India’s government has recently made a commitment to contribute a larger portion of its gross domestic product (GDP) to the healthcare system, but this funding requires efficient and effective structures to make a significant impact. The objective of this study is to measure the technical efficiency (TE) and scale efficiency (SE) of several government-run primary healthcare centers (PHCs) to determine whether changes in structure might be indicated for more efficient resource allocation. This study uses DEA, and related methods to analyze the efficiency of 39 government PHCs in Southern Karnataka, India. Three inputs (clinical staff, non-clinical staff, and population coverage) and five outputs (outpatient visits, institutional deliveries, children immunized, antenatal care visits, and family planning visits) were used in the analysis. A Tobit censored regression model was conducted as a second-stage analysis. Fourteen (36%) of the 39 PHCs analyzed were found to be efficient assuming constant returns to scale (CRS), with an average efficiency of 84%. Twenty one (54%) of 39 were found to be technically efficient assuming variable returns to scale (VRS), with an average efficiency of 90%. Fifteen (38%) of the 39 PHCs displayed scale efficiency (SE), with an average SE of 93%. Stochastic Frontier Analysis (SFA) showed that the majority of the variance in the samples was due to inefficiency rather than noise in the data, and was highly correlated (.989) with the DEA model. Given these results, inefficient PHCs could hypothetically decrease their inputs by an average of 10% to achieve pure VRS technical efficiency and 7% to achieve scale efficiency. Because a large number of the PHCs are operating at increasing returns to scale, the additional financial resources slated for the public healthcare system would likely help the PHCs to become more efficient. This could increase quality and access to healthcare for many people of India. The authors suggest further study of additional PHCs in Karnataka to foster exploration of system- wide efficiencies and inefficiencies. Furthermore, implementation of a periodic monitoring of the PHCs over time as well as a qualitative evaluation of the health care system would provide additional insights.
Name: Charlotte Marous MS '12
Degree and Year: Health and Exercise Science
Wake Forest University, 2011
Capstone Project: Prevalence, risk factors, and clinical epidemiology of hepatitis C; an emerging infectious disease in the Alor Setar region of Kedah, Malaysia
Abstract: Hepatitis C Virus (HCV), a chronic infectious disease caused by a small single-stranded RNA virus (family Flaviviridae) and major contributor to liver disease worldwide, is believed to be an emerging global threat. However, data on the prevalence, risk factors, and demographic epidemiology of the disease in the Kedah state of Malaysia are still limited. We carried out a hospital-based retrospective analysis of 707 confirmed cases extracted from a Hepatitis C Registry and past patient medical records set up from 2009 in Hospital Sultanah Bahiyah, Alor Setar, Kedah state, Malaysia, with the goal of providing a primary analytic description of the contemporary prevalence, risk factors, and clinical epidemiology of the disease in this potentially high risk region of the country.
The mean age of patients was 41.64 years (range: 1-104 years), with infection varying nonlinearly with age and gender. Males (88.0%) were the predominant gender affected by HCV and the majority of cases were Malays (79.5%). Intravenous drug use (64.2%) was the overwhelming prime risk factor associated with infected individuals, with incarceration(11.9%) and blood transfusions(8.6%) following in second and third, respectively. Diabetes mellitus (10.0%) and hypertension (9.1%) were the largest associated chronic diseases found in conjunction with HCV positive cases. Both HCV acute infection in addition to chronic liver cirrhosis and hepatocellular carcinoma seemed to follow a bell-curve prevalence pattern with respect to age; the largest prevalence topping off around age 40-50 years. This trend persisted when adjusted for gender as well. HCV appears to be strongly related with middle age, Malay ethnicity, and male gender, along with associated chronic diseases including diabetes mellitus and hypertension. Furthermore, intravenous drug use, incarceration, history of blood transfusion, and sexual promiscuity seemed to be the most common risk factors playing a role in infection rates. Incomplete medical records inhibited economic, treatment, and outcome of HCV to be evaluated in this particular study population so further investigation is necessary to estimate both current and future disease burdens in the state of Kedah. This data provides initial insight into the major demographic trends, populations sectors, and most common risk factors associated with HCV infections in this potentially risky area of Malaysia.
Name: Anna Nanigian MS '12
Degree and Year: Preprofessional Studies and Sociology
University of Notre Dame, 2011
Capstone Project: Does green always mean 'go'?: Assessing the extent to which India’s National Rural Health Mission traffic light monitoring and evaluation system accurately reflects the performance of Primary Health Centers in the state of Karnataka
Abstract: In the state of Karnataka, an evaluation and planning initiative termed the Community Planning and Monitoring of Health Systems (CPMHS) has been formed to integrate constituent participation into health care reformation. In conducting CPMHS, representatives of each village, called Village Health and Sanitation Committee (VHSC) members assess the success of their Primary Health Centers (PHCs) by rating eleven different health parameters, such as maternal health, child health, disease surveillance, quality of care, and others, as red, yellow, or green, on a quarterly report card. An analysis of whether these traffic light evaluations accurately reflect the performance of PHCs, however, has not yet been undertaken. The goal of the research was to assess the extent to which traffic light scoring of PHCs is congruent with those PHCs’ ability to adequately service their constituents. In addition to performing quantitative analysis, the second aim of this research project was to analyze the social processes which take place to formulate traffic light rankings in community dialogue sessions. In my research, I found that from 2010 to 2011, the highest rated health parameters, those with the most villages reporting “green”, were Maternal Health Guarantees, Child Health, and Community Perceptions of ASHA. This identified the PHC goals, but did not relate them back to epidemiological data. Though a primary goal of this project was not achieved, in that I was unable to fully compare traffic light data to objective epidemiological data from PHC to PHC, the data I was able to collect, as well as the gaps in those data, shed light on the efficacy and benefits of the process of community monitoring, as well as the shortcomings which hamper it from being a truly effective method of accountability within the government health care system.
Name: Christopher Wynkoop MS '12
Degree and Year: Chemistry
University of Notre Dame, 2011
Capstone Project: Capturing cholera: Understanding disease dynamics and seasonal variations in Kolkata, India via a mathematical model
Abstract: Understanding the complex dynamics of cholera can help to develop intervention strategies for current epidemics and predict and possibly prevent future outbreaks. In our research we developed a mathematical model that takes into account intrinsic disease dynamics, the presence of an aquatic reservoir of Vibrio cholerae, and external seasonal forcing factors for the Kolkata area in India. We performed a stability analysis to determine an expression for the time-averaged expression of R0, the basic reproduction ratio, and find that it underestimates the true value of R0. We also performed a sensitivity analysis and found several parameters that most influence the model output and, therefore, disease dynamics. We combined the aforementioned components of our model into one SIRS-type model to account for the intrinsic dynamics of a cholera outbreak determined by population-driven parameters and a seasonality component driven by external climate forcing factors. The parameter that has the strongest effect on disease dynamics is the recovery rate; there is a significant reduction in both the total size and severity of an epidemic when this rate is higher in a population. We will continue this project with a resonance analysis and testing of the model to ensure it is ready for use with cholera and climate data in Kolkata.
Name: Amy Billow MS '13
Degree and Year: Biology
University of Cincinnati, 2012
Capstone Project: The impact of the diabetes epidemic in India: Establishing the prevalence and clinical profile of Metabolic Syndrome in Type 1 Diabetes patients in Chennai, India
Abstract: The lifestyle shift that is taking place in India has introduced type 2 diabetes mellitus (T2DM), obesity and the metabolic syndrome (MS) in epidemic proportions. While the connection between MS and TD2M is well studied, the connection between type 1 diabetes mellitus (T1DM) and MS is less understood. Although lifestyle choices do not influence onset of T1DM, increasing obesity may be a key player in the evolution of insulin resistance and “double diabetes.” This study sought to establish the prevalence of MS in T1DM and investigate how presence of MS affects prevalence of diabetes complications using a T1DM cohort at Dr. Mohan’s Diabetes Specialty Center in Chennai, India. A retrospective analysis using an electronic medical record system was done on 491 T1DM patients. Relevant clinical and biochemical data from patient first visit was collected and MS was diagnosed according to International Diabetes Federation (IDF) criteria. MS risk factors and macrovascular and microvascular complications data was analyzed for association with MS using t-test, chi-square and logistic regression.
Prevalence of MS was 7.9% overall. MS positive people had longer duration of T1DM, were overweight or obese, showed a family history of DM, showed signs of insulin resistance and had higher cholesterol. Presence of MS was not associated with poor glycemic control. MS positive patients had higher prevalence of microvascular complications but not macrovascular complications, with a strong association between MS and presence of nephropathy even after adjusting for confounding variables. In general, the prevalence of microvascular but not macrovascular complications rose with number of MS components present. This study has established a baseline prevalence of MS in a T1DM cohort in Southern India. There is a clear association between certain risk factors such as duration of DM and obesity and presence of metabolic syndrome. Presence of MS at first visit increases the odds that patients have microvascular complications such as retinopathy, neuropathy and especially nephropathy at first visit, although further study is needed to determine a causal relationship between presence of MS and development of complications.
“I was able to learn what it takes to carry out a Global Health research project from start to finish. In doing this, I was able to improve upon my knowledge of a particular disease, adjust to cultural differences in a research setting, and improve upon both my analytical and my writing skills.”
Name: Brian Bush MS '13
Degree and Year: Biology
University of Notre Dame, 2012
Capstone Project: 5 year (2008 -2012) retrospective analysis of the National Cancer Patient Registry - Colorectal Cancer Database in Kedah, Perlis, and Seberang Jaya, Malaysia
Abstract: Assuming the validity of association between economic/human development and colorectal cancer, Malaysia is an especially important country to examine, given its recent developmental trajectory and commitment to development. Given the research demonstrating the lack of awareness among Malaysians about colorectal cancer, disease prevention and benefits of screening/early detection, it behooves both health care providers and the public at large to have accurate and up-to-date information about colorectal cancer on a longitudinal scale. The long-term goals of the research project are to assess colorectal cancer trends retrospectively on a national level. We gathered a population of colorectal cancer patients and determined specific causes of each case and symptoms and signs that can present in the disease. The results of this study suggest that males potentially are at more risk than females. The data does not provide conclusive evidence that any one ethnic group in the states of Kedah, Perlis, and Seberang Jaya are more at risk for colorectal cancer than any other. However the data suggests that age is one of the most significant risk factors associated with colorectal cancer. Ideally, the retrospective data obtained from this project may provide a sound basis for public health officials and heath care workers, as they develop new ways to publically address and treat colorectal cancer in Malaysia
“Conducting research abroad afforded me the chance to make a meaningful contribution to helping a serious health problem, colorectal cancer, in a country of need. Moreover, the independence that I had on this project was a great asset that has helped me to mature as a person and scientist.”
Name: Macy Brusich MS '14
Degree and Year: Integrative Biology and Environmental Science
University of Illinois- Urbana Champaign, 2012
Capstone Project: Targeting educational campaigns for prevention of vector-borne disease: An assessment of rural vs. urban settings in Thailand
Abstract: The goal of this study was to collect information that will help to reduce the risk of malaria and/or dengue fever diseases in at-risk populations. The study utilized a mixed method design to capture data using a forced choice and open-ended questionnaire that assessed household construction and knowledge, attitudes, and practices (KAP) of patients seeking point-of-care treatment for malaria and/or dengue fever in a rural and urban district of Thailand. Additional quantitative findings from household mosquito collections were also performed. Study objectives were to compare individual and household level health practices and to quantify differences in KAP between rural and urban settings. Sixty-four participants were recruited in total to assess differences in key variables at the individual and household level that may influence health behaviors related to the prevention of malaria and dengue fever disease. While statistically significant findings should be interpreted with caution, results from this pilot study indicate knowledge surrounding malaria and dengue fever exists, however gaps in behavior, perception, and adequate protection from mosquito vectors were found. Significant associations between study site and household construction were found to influence household level practices in the types of mosquito control products purchased and the abundance of mosquitoes in homes. Overall, education from malaria and dengue fever Ministry of Health intervention campaigns is reaching the intended target populations. Results are intended to guide future MOH health education campaigns in these study settings that target specific community needs.
“This field experience allowed me to gain transferrable skills that I can apply in future global health settings as I begin my exciting career path in Global Health.”
Name: Christiane Shizuko Cardoza MS '14
Degree and Year: Environmental Sciences
University of Notre Dame, 2010
Capstone Project: Non-partner sexual violence against women and children in Gujarat, India
Abstract: Non-partner sexual violence is a severe but undocumented problem within India as a whole and within Gujarat specifically. Sensationalized events have occurred in the past decade, bringing the spotlight to events of non-partner sexual violence. The purpose of this paper was to create a methodology for measuring and tracking non-partner sexual violence, to use this methodology to begin to evaluate non-partner sexual violence in India, to understand determinants of cyclical sexual violence, and to propose solutions to address both the problem of event monitoring and the determinants of non-partner sexual violence. A content analysis of Indian-based English-language newspapers and five informational interviews was performed. Although events data were not available for each month, this study demonstrated how aggregate levels of non-partner sexual violence could be illustrated and compared. In order to address sexual violence events, several interventions are recommended including continued and improved events monitoring, crowd-sourcing and crowd-feeding, and education and empowerment programs.
“Using data we gathered, Will developed a platform to map sexual violence event reports from eyewitnesses through texts on a basic cell phone, providing another option for victims to report these crimes.”
Name: Rebecca Tracy MS '14
Degree and Year: Psychology and Pre-Health Studies
University of Notre Dame, 2013
Capstone Project: Post-project sustainability study of the Child Health Opportunities Integrated with Community Empowerment (CHOICE) Project, Indonesia: correlations between cognitive and physical development and infant nutrition
Abstract: Post-project sustainability studies allow for an analysis of the outcomes and the long-term benefits of a development programs. In this study Project Concern International’s (PCI) Child Health Opportunities Integrated with Community Engagement (CHOICE) program is evaluated seven years after its completion for its long term affects on children’s cognitive development. A survey of CATCH indicators was utilized along with an infant development checklist and the Raven’s Coloured
Progressive Matrices for data collection. Two populations (0-2 year olds and 8-10 year olds) were evaluated in order to look at the effects the CHOICE program had on development both from when the program was run and its lasting effects on the current population. The development of 0-2 year olds and 8-10 year olds was modeled using general linear modeling and analyzed based on different areas affected by CHOICE program activities in particular children’s nutrition and physical growth. While there were no overall significant differences found between CHOICE villages and control villages due to treatment, trends in the data suggest that CHOICE villages performed better on the development tests compared to control villages as well as on many of the outcome indicators the CHOICE program aimed to improve. Based on this analysis the CHOICE program activities related to nutrition, hygiene and maternal knowledge made an impact on the community, which was sustained seven years later. Based on the research presented in this paper development programs that aim to improve the nutritional status of mothers and children can have long-term positive effects on children’s cognitive development.
“My specific portion of this study was focused on one of the long-term aspects of nutrition and health improvement programs – cognitive development.”
Name: Thomas Ulsby MS '14
Degree and Year: Biology
Gustuvus Adolphus College, 2012
Capstone Project: The impact of globalization on nutrition, KAP, and health of type II diabetics at Dr. Mohan’s Diabetes Specialties Center
Abstract: Diabetes is now considered a global pandemic showing particularly severe effects in low and middle-income populations. Although all populations are affected, depending on a person’s genetic background, family history, health, and behavior, the susceptibly for the disease may rise or fall. Health, behavior, and knowledge impact the attention a person will devote to health challenges and changes in their life which may in turn be impacted by globalization. The current study aimed to investigate if rural-urban differences in these variables impact the spread of Diabetes Mellitus in India via the carrying out a comparative case study in two Indian population settings. The project investigated differences in 24 hour nutrition recall and Knowledge, Attitudes, and Practices (KAP) of type II diabetic patients from Dr. Mohan’s Diabetes Specialties Center at Gopalapuram, Chennai, Tamil Nadu (located in a major Indian urban setting) with Chunampet, Kanchipuram, Tamil Nadu (located in a more rural setting). 225 outpatients were surveyed in total (110 from Gopalapuram and 115 from Chunampet). 15 survey respondents (5 from Gopalapuram and 10 from Chunampet) were then interviewed to provide contextualization for the trends represented in the summarized survey. Respondent answers were scored by variable, analyzed, and summarized. Differences in nutrition and knowledge scores were found between the study populations (< 0.05 = 95%). Nutrition and knowledge scores produced negative regressions when controlling for location (< 0.05 = 95%), suggesting that rural variables reduce nutrition and knowledge scores of diabetic patients. However, even with these differences diabetes, is still prevalent in both locations. Globalization may be influencing the opportunities of type II diabetes patients for practicing good nutrition and health. As interactions between urban and rural centers grow, the spread of more urban and non-communicable health problems will also spread into the rural environments. As this research shows, current differences in nutrition and knowledge between urban and rural centers may indicate an association with the presence of a disease and poor care experienced by rural patients. However, despite this finding, the fact that diabetes is still highly prevalent among the urban subjects indicates that other factors may need to be evaluated to explain the observed disease variability. Health behavior models, Integral Human Development, and economic growth may help explain why these differences were observed and what can be done to change the epidemic.
“It has been a rewarding experience and I have felt extremely blessed to be here among such great researchers and doctors. Thank you for your help, guidance, hospitality, and care.”