Adam T. PerzynskiApril 1, 2008<br />Between Facts and Voices:Medical and Lay Knowledge of the Spread of Hepatitis C<br />E...
Committee<br />Dale Dannefer, Ph.D.<br />Gary Deimling, Ph.D.<br />Susan Hinze, Ph.D.<br />Neal Dawson, M.D.<br />
Between Facts and Voices<br />Social Fact<br />	Here then, is a category of facts with very distinctive characteristics: i...
Introduction<br />Hepatitis C is a widespread and serious disease that affects the liver. <br />170 million people worldwi...
HCV Transmission<br />Blood <br />Intravenous Drug Use<br />Blood Transfusions<br />Needle Sticks <br />Shared Household I...
Personal Significance<br />Narratives of Liver Transplantation: Recipient Perspectives on the Quality of Life<br />Researc...
Theoretical Significance<br />Variation in what lay people know is most often assumed to be a matter of ignorance and misi...
Practical Significance<br />Most people with HCV are experiencing no symptoms and are completely unaware that they are inf...
Research Objectives<br />To understand the forms and distribution of knowledge of the spread of HCV.  <br />To describe pa...
Analysis Objectives and Contributions<br />Assess the level of knowledge of HCV transmission.<br />Describe and explain ra...
Why Study Knowledge?<br />Knowledge is associated with health behavior and health.<br />Most investigations of lay people’...
Figure 1: Knowledge of the Spread of HCV<br />
Physician Knowledge<br />38% of physicians surveyed by the American Gastroenterological Association (AGA) believed that th...
Theoretical Frameworks for Understanding Illness Knowledge<br />Sociology of Knowledge<br />Knowledge Diffusion<br />Knowl...
A Mixed Methods Approach<br />Two Data Sources<br />Qualitative Primary Data<br />Quantitative Secondary Data<br />Crystal...
Qualitative Interviews<br />42 in-depth interviews with HCV patients <br />Phase I of Alcohol Reduction in Medical Illness...
Qualitative Interviews N = 42<br />38% Female<br />38% Black, 36% White, and 26% Hispanic<br />Mean time since diagnosis w...
Qualitative Analysis Techniques<br />Incremental Team Based Approach<br />Open coding by hand on paper.  <br />Generate th...
Qualitative Analysis Techniques<br />Axial Coding (Stauss & Corbin 1990)<br />Choose specific themes<br />Spread of HCV<br...
Quantitative Sample<br />Behavior Risk Factor Surveillance System (BRFSS), 2001, Arizona<br />Conducted by the Centers for...
Measures in the CDC Survey<br />Sociodemographic Characteristics: Age, Sex, Race, Income and Education<br />Do you know so...
Measures of the Knowledge of the Spread of HCV<br />Do you think hepatitis C can be spread thru?<br />Sneezing or Coughing...
Methods of Quantitative Analysis<br />Managed with SPSS<br />Analyzed with MPlus<br />Analysis proceeded in several stages...
Qualitative Results<br />Where Patients Talked About HCV<br />How they themselves were infected with HCV<br />RJ:  That pa...
Qualitative Results<br />What Patients Said<br />How they themselves were infected with HCV<br />RJ:  That part, I’m prett...
Qualitative Results<br />	“I don’t wear a sign saying, hey I got hepatitis C, you know, don’t use the bathroom. You know, ...
A Qualitative Example<br />	“One doctor told me they’ve discovered cases in people they figured they contracted back in th...
Distribution of Outcome Variables<br />
Description of Sample<br /><ul><li>  Mean Age was 48.5 (SD 18.1)
  59% Female
  22% Know someone with HCV
  25% Said they had been tested for HCV
  75.8% White, 17.5% Hispanic, 4.4% Native American / American Indian, 2.3% Black
  Mean Income Level was $50000-$75,000
  Median Education Level was completion of high school</li></li></ul><li>Exploratory Factor Analysis<br />Unweighted Least...
Exploratory Factor Analysis [CONSIDER SHOWING PROMAX]<br />EXPLORATORY ANALYSIS WITH <br />1 FACTOR :<br />ROOT MEAN SQUAR...
CHECK DF FOR ACCURACY <br />That is probably a typo<br />
Fix spelling error above<br />
Three Latent Classes<br />2,4,5,6, and 7 category models do not fit the data as well as the 3 category model. <br />
Figure 4: Distribution of Forms of Knowledge (Most Likely Latent Class)<br />
Figure 5: Estimated Probabilities of Knowing How HCV is Spread by Most Likely Latent Class<br />
Structural Model Testing<br />What concepts are associated with the likelihood of each form of knowledge?<br />Mixture Mod...
Results of Structural Model Testing<br />Strongest Associations<br />Race/ethnicity <br />Knowing someone with HCV<br />Fe...
Effects of Covariates on Form of Knowledge: Everywhere vs. True Awareness<br />N = 3092<br />
Effects of Covariates on Form of Knowledge: Nowhere vs. True Awareness<br />N = 3092<br />
Probability of Form of Knowledge by Race/Ethnicity (N = 3092)<br />
Probability of Form of Knowledge by Knowing Someone w/ HCV (N = 3092)<br />
Probability of Form of Knowledge by Education Level<br />
Probability of Form of Knowledge by Age<br />
Probability of Form of Knowledge by Whether R Had a Health Plan (N = 3092)<br />
Discussion<br />Most important excerpts from Ch. 7<br />Summary of surprising findings<br />
Limitations<br />Secondary Data<br />Limited Items about HCV <br />Arizona<br />Confidence in / certainty of knowledge<br ...
Upcoming SlideShare
Loading in …5
×

Between Facts and Voices: Medical and Lay Knowledge of the Spread of Hepatitis C

860 views
763 views

Published on

Dissertation Defense slides for Adam T. Perzynski, PhD

Defended in 2008 at Department of Sociology, Case Western Reserve University

Published in: Education, Health & Medicine
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
860
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • Most social scientific studies of health, illness, and health behavior assume that lay people’s knowledge can be adequately described as a matter of ignorance and misinformation judged against the paradigm of biomedical, scientific knowledge (Abraham et al 2000). However, research on lay people’s actual understanding and knowledge of how HCV is spread is scarce (Hopwood & Southgate 2003). The little research that has been conducted on lay knowledge of hepatitis C is a patchwork of research methods and perspectives. Studies of illness knowledge most often incorporate models from health psychology and epidemiology, whereby lay people’s knowledge is evaluated based upon its correspondence to accepted scientific standards (Brown 1995;).
  • Most social scientific studies of health, illness, and health behavior assume that lay people’s knowledge can be adequately described as a matter of ignorance and misinformation judged against the paradigm of biomedical, scientific knowledge (Abraham et al 2000). However, research on lay people’s actual understanding and knowledge of how HCV is spread is scarce (Hopwood & Southgate 2003). The little research that has been conducted on lay knowledge of hepatitis C is a patchwork of research methods and perspectives. Studies of illness knowledge most often incorporate models from health psychology and epidemiology, whereby lay people’s knowledge is evaluated based upon its correspondence to accepted scientific standards (Brown 1995;).
  • The outcome variable in this research is knowledge of the spread of HCV, but whose knowledge are we talking about? What are the possible sources of that knowledge? How is lay people’s knowledge related to scientific knowledge? In this research I do not assume that scientific knowledge is universally more factual or informative than knowledge with cultural origins other than science. Instead I treat science and culture as primary knowledge sources. Figure 1 illustrates two broad sources of “Knowledge of the Spread of Hepatitis C” (KSHC). The diagram shows that the knowledge banks of healthcare providers, HCV patients, and lay people are all linked to science and culture. In my own position as a social scientist, I aim to describe not only how I. Sociocultural Knowledge matches up with II. Scientific Knowledge but also the cultural forms that KSHC (knowledge of the spread of HCV) takes and the patterns of interactions that occur between groups of people with different forms of KSHC. Note that the paths and concepts in Figure 1 should not be seen as either mutually exclusive or exhaustive. For example, there are many nurses infected with HCV. These nurses would fit into both the III. Provider and V. Patient boxes. Figure 1 is a visual tool for understanding the arrangement of knowledge of the spread of HCV in this dissertation. In the following sections I describe the numbered components of Figure 1.
  • The findings from a study of the HCV knowledge of primary care residents are somewhat alarming. The sample of 180 primary care residents at five U.S. training programs found that the residents were unaware of the appropriate risk factors that indicate a patient should be tested for HCV. Most surprising is that 66% recommended vaccination for HCV, even though there is no vaccine for HCV (Coppola, Karakousis, Metz, Go, Mhokashi, Howden, Raufman and Sharma, 2004).
  • Wallerstein and Smelser (1969) argued that qualitative and quantitative research methods could be used to create a “complementary articulation” of research findings. While this project does not follow the same procedures as suggested by Wallerstein and Smelser, the idea of complementary articulation permeates my research. In particular, I use qualitative findings to solidify ideas, formulate research questions, specify and re-specify models, and interpret quantitative results. Complementary articulation does not really refer to the overall design of a study. Instead, it points out that the knowledge learned from qualitative and quantitative findings will be used to better describe aspects of the same reality (Smith 2004).
  • Patients had to have had one drink in their lifetime and not be a current alcohol abuser. Patients who had recovered from HCV were ineligible.
  • 16 Women and 26 Men16 Blacks 15 Whites 11 Hispanics
  • The analysis of the qualitative interview data was an incremental process. First, open coding as described by Strauss and Corbin (1990) was conducted by a team of four qualitative researchers by hand on paper. Open coding was directed at generating an initial list of themes and codes representing patterns in the data. Regular meetings were held to discuss the coding and develop a preliminary list of codes. The list was elaborated with continued coding. Once a solid initial coding scheme had been developed, paper coding was transferred to NVivo, and coding continued exclusively using the software. The research team divided the coding equally. Most interviews were coded by a single researcher. Periodic checks of interobserver consistency were instituted. All of the researchers would regularly code a single interview and meet to make a thorough comparison.
  • In the next step in the analytic process, the coding was much like the axial coding described by Strauss and Corbin (1990). I chose specific themes that matched with the research questions for this dissertation. In particular, I focused on one specific category, the Spread of HCV as described by the patients in the interviews. While grounded in the words of the patients, my approach to this stage of coding diverges from Strauss and Corbin in that it was directed at comparing patient perspectives on the spread of HCV with the operationalization of KSHC that was used in the quantitative part of the study. Therefore while the structure of the coding was partly determined by being grounded in observations from the interviews, it was also partly determined by a structure that I imposed on the data. In order to complement the quantitative analysis, a section of the results of the qualitative coding presented in Chapter 4 are presented in relation to their presence or absence the CDC survey. It should also be noted that my own subjective lens and technique for handling the nuts and bolts of the textual coding process borrows from Michel Foucault. Foucault suggests that researchers look for the meaning of texts by starting at the surface rather than delving into deep conceptual analyses. Foucault referred to this approach the “archaeology of the text” (Foucault 1961). In archaeology, workers at the field dig site start on the surface, with an initial survey. Pieces of the field site are then marked off into what are typically two meter squares. Each square is assigned to an individual field technician, or a series of technicians. The job of the technician is to scrape off no more than 1/16 of an inch of soil with each pass of the trowel. Any artifact discovered in each layer of soil is immediately charted and described in detail in a graph paper journal. My approach to the qualitative data is somewhat of an “archaeology of the text.” While the interviews contain topic on hundreds of diverse themes of interest, my two meter square is Knowledge of the Spread of HCV. In looking at the text, I focused almost exclusively on sections (or textual artifacts) where patients described their knowledge of the spread of HCV, or knowledge that others had. I started on the surface, first looking at where people discussed knowledge of the spread of HCV, before considering what they said. With each subsequent pass of my highlighter I learned a little bit more about the data.
  • Insert text RE: race here from Ch. 2
  • LMR Test
  • It’s important that there be a substantive/conceptual fit.
  • First, the quantitative data are secondary data. I did not participate in the collection of the CDC data and I was therefore not involved in the selection of the questions to be included in the HCV module. Upon review of the qualitative results, it is apparent that at least among HCV patients, lay people have broad understandings of how HCV might be spread. The range of these understandings (or often misunderstandings) is not matched by the range of items administered by the CDC in Arizona. I do not know how the individuals in the Arizona CDC sample would have responded to questions about whether HCV could be spread by swimming with someone, tasting someone’s blood, or getting tattoos. There exists a possibility, that given a broader range of survey questions, a fourth or a fifth form of knowledge could be identified, or that some individuals’ likelihood of a particular form of knowledge (most probable latent class) could change. The second fundamental limitation is that the data are from a single state, Arizona, which is culturally distinct from other states in America. Thus these findings, even the appearance of three distinct forms of knowledge of the spread of HCV, should only be specifically generalized to the population of the state of Arizona. Further, since some racial and ethnic groups, like Asian Americans, were not included in the analysis, my conclusions may not apply to those groups. I am excited about the possibility that these forms of KSHC, and that illness knowledge more generally may exist in several forms in society, but I cannot make that determination based solely upon the data in this research project. Fourth, I did not weight the data from the CDC in Arizona. Although the CDC’s BRFSS Manual provides a formula for weighting the data, the data analysis software (MPLUS) was not capable of the computations necessary for the properly weighting of the data at the time of the analysis. Finally, in the latent class analysis with covariates (Chapter 6), it was not possible to test a model that accounts for the relationships among the independent variables. In a path analysis, researchers are able to test direct and indirect effects of the variables in the model. For example, gender and education may have an important relationship with income which then has an effect on knowledge. These indirect effects could not be tested using latent class analysis with covariates because of limitations in MPlus, the mixture modeling software.
  • The findings demonstrate that Classical-Test Theory, or the quiz approach, is not always a valid representation of how illness knowledge varies. Assuming that knowledge only takes one form can lead to the misspecification of models and invalid and unreliable conclusions. I did not choose latent class analysis because it was a familiar statistical technique, or because I wanted to demonstrate my expertise with a new an advanced software package. I chose LCA because I did not assume that other techniques were telling the whole story about knowledge of the spread of HCV. The mixed methods design of this research deserves much of the credit for challenging how we think about the forms and distribution of illness knowledge. I was only prepared to formulate a new conceptualization of KSHC because I had taken the time to examine the textual ethnographic details of the knowledge and experiences of HCV patients. Based on the percentage of people who have the Nowhere form of knowledge, it is reasonable to conclude that without a massive public health education program of the same scale as HIV/AIDS education programs, most people who currently have HCV will continue to be unaware of their infection until they have symptoms. These symptoms are typically associated with more advanced stages of chronic HCV infection at which point existing treatments are increasingly less effective. Campaigns to educate people about how HCV is spread should take into account the fact that some people think HCV is everywhere, some people think HCV is nowhere and some people have a good idea of how you can and cannot get HCV. The risks of ignoring different forms of KSHC are far from minimal. A recent public health advertisement in Cleveland, Ohio at a popular sports venue visited weekly by tens if not hundreds of thousands of people, displayed clearly that HCV could be spread by “household items.” This type of information is far from helpful to anyone, and could be potentially harmful if we are to assume that a group of people in society already think that you can get HCV from “everything.” What the billboard meant to say was that household items like nail clippers, razors, and toothbrushes can be contaminated with infected blood and that these specific items are known to transmit HCV.
  • These findings are also important for evaluating the “knowledge diffusion” model advanced by Pampel (2002). In the knowledge diffusion model, innovations, such as scientific knowledge of how HCV is spread, are diffused from high status groups where they originate downward to through the rest of society by order of the status hierarchy. There is no evidence that high status groups were the first to come up with the Everywhere form of KSHC. In fact, it is far more likely that this form of knowledge, in which HCV can be spread by “anything under the sun,” originated with the low status groups among which it is the most common. Knowing someone with HCV had a strong association with having the True Awareness form of knowledge, while having health insurance, having had an HCV test, and having a blood transfusion were not positively associated with having the True Awareness form. This suggests that KSHC, and potentially other types and sets of illness knowledge are diffused more efficiently through lay networks than through the healthcare system.
  • These findings are also important for evaluating the “knowledge diffusion” model advanced by Pampel (2002). In the knowledge diffusion model, innovations, such as scientific knowledge of how HCV is spread, are diffused from high status groups where they originate downward to through the rest of society by order of the status hierarchy. There is no evidence that high status groups were the first to come up with the Everywhere form of KSHC. In fact, it is far more likely that this form of knowledge, in which HCV can be spread by “anything under the sun,” originated with the low status groups among which it is the most common. Knowing someone with HCV had a strong association with having the True Awareness form of knowledge, while having health insurance, having had an HCV test, and having a blood transfusion were not positively associated with having the True Awareness form. This suggests that KSHC, and potentially other types and sets of illness knowledge are diffused more efficiently through lay networks than through the healthcare system.
  • Between Facts and Voices: Medical and Lay Knowledge of the Spread of Hepatitis C

    1. 1. Adam T. PerzynskiApril 1, 2008<br />Between Facts and Voices:Medical and Lay Knowledge of the Spread of Hepatitis C<br />E-mail: Adam.Perzynski@case.edu<br />This research was supported in part by NIH Grant No. 1 R01 AA13302-01A1<br />
    2. 2. Committee<br />Dale Dannefer, Ph.D.<br />Gary Deimling, Ph.D.<br />Susan Hinze, Ph.D.<br />Neal Dawson, M.D.<br />
    3. 3. Between Facts and Voices<br />Social Fact<br /> Here then, is a category of facts with very distinctive characteristics: it consists of ways of acting, thinking, and feeling, external to the individual, and endured with a power of coercion, by reason of which they control him.<br />Emile Durkheim (1895), Rules of the Sociological Method pp. 2-3. <br />Voice<br />4 a : wish, choice, or opinion openly or formally expressed &lt;the voice of the people&gt; b : right of expression; also : influential power<br />From Merriam-Webster (2003)<br />
    4. 4. Introduction<br />Hepatitis C is a widespread and serious disease that affects the liver. <br />170 million people worldwide are infected with the Hepatitis C virus (HCV).<br />4.1 million Americans infected with HCV. (CDC 2007) <br />As many as 10,000 Americans die every year from chronic HCV infection.<br />Symptoms of chronic HCV include fatigue, loss of appetite, jaundice, itching, and depression. (CDC 2004) <br />
    5. 5. HCV Transmission<br />Blood <br />Intravenous Drug Use<br />Blood Transfusions<br />Needle Sticks <br />Shared Household Items (Razor or Toothbrush) <br />Sexual transmission of HCV is recognized but is infrequent. <br />HCV is NOT transmitted by Coughing, Kissing, Sneezing, Touching, Bathrooms, Fecal Matter, or Contaminated Food<br />
    6. 6. Personal Significance<br />Narratives of Liver Transplantation: Recipient Perspectives on the Quality of Life<br />Research on lay knowledge of how HCV is spread is scarce (Hopwood & Southgate 2003). <br />
    7. 7. Theoretical Significance<br />Variation in what lay people know is most often assumed to be a matter of ignorance and misinformation as judged by biomedical, scientific standards (Abraham et al 2000; Brown 1995). <br />Research on lay knowledge of how HCV is spread is scarce (Hopwood & Southgate 2003). <br />
    8. 8. Practical Significance<br />Most people with HCV are experiencing no symptoms and are completely unaware that they are infected with the disease. (CDC 2007)<br />Testing for HCV relies on identification of transmission risks (Gordon 1999). <br />
    9. 9. Research Objectives<br />To understand the forms and distribution of knowledge of the spread of HCV. <br />To describe patterns of interaction between people with different forms of HCV knowledge. <br />To test a model of sociocultural predictors of knowledge of the spread of HCV. These predictors include race, ethnicity, gender, and socioeconomic status. <br />
    10. 10. Analysis Objectives and Contributions<br />Assess the level of knowledge of HCV transmission.<br />Describe and explain racial and ethnic variation in lay knowledge and understanding of HCV. <br />Develop a measurement model for knowledge of HCV transmission. <br />Test a causal model of what predicts knowledge of HCV transmission.<br />
    11. 11. Why Study Knowledge?<br />Knowledge is associated with health behavior and health.<br />Most investigations of lay people’s health or illness knowledge assume a knowledge deficit.<br />Low score on Knowledge Quiz = Knowledge Deficit<br />Most health care is self care. <br />
    12. 12. Figure 1: Knowledge of the Spread of HCV<br />
    13. 13. Physician Knowledge<br />38% of physicians surveyed by the American Gastroenterological Association (AGA) believed that their HCV patients should not hold jobs in food preparation (Harris Interactive 2003, p. 38). <br />35% of primary care physicians, and 40% of gastroenterologists surveyed by the AGA agreed that if they had hepatitis C they would not want anyone to find out (Harris Interactive 2003, p. 48). <br />Another study of specialists (N=1,249) found that 25% recommended that their patients not share drinking glasses or dishes (Everhart, 1997). <br />
    14. 14. Theoretical Frameworks for Understanding Illness Knowledge<br />Sociology of Knowledge<br />Knowledge Diffusion<br />Knowledge Gap<br />Social Inequality<br />Cumulative Advantage<br />Culture and Structure<br />
    15. 15. A Mixed Methods Approach<br />Two Data Sources<br />Qualitative Primary Data<br />Quantitative Secondary Data<br />Crystallization of theories and Ideas<br />Complementary Articulation (Smelser & Wallerstein)<br />Inductive and Deductive<br />
    16. 16. Qualitative Interviews<br />42 in-depth interviews with HCV patients <br />Phase I of Alcohol Reduction in Medical Illnesses (ARIMI Study)<br />Conducted from January 2002 to August 2004. <br />Face-to-face interviews lasted 30-90 minutes<br />Tape-recorded and transcribed verbatim<br />Urban Community Medical Center<br />Recruitment<br />In person at a gastroenterology clinic <br />By phone with people who had an HCV diagnosis in the emergency department but did not follow-up with a doctor afterward. <br />
    17. 17. Qualitative Interviews N = 42<br />38% Female<br />38% Black, 36% White, and 26% Hispanic<br />Mean time since diagnosis with HCV was 8 years (SD=6.5)<br />Mean education of 12 years (SD= 2.3)<br />Age ranged from 36-74 (mean=49.5, SD=8.1). <br />
    18. 18. Qualitative Analysis Techniques<br />Incremental Team Based Approach<br />Open coding by hand on paper. <br />Generate themes<br />Observe patterns<br />Regular meetings to discuss the coding and develop a preliminary list of codes. <br />Paper coding was transferred to NVivo qualitative data analysis software<br />Coding was divided among the team equally<br />Periodic checks of interobserver consistency<br />
    19. 19. Qualitative Analysis Techniques<br />Axial Coding (Stauss & Corbin 1990)<br />Choose specific themes<br />Spread of HCV<br />Fears<br />Transmit to Others<br />The “archaeology of the text” (Foucault 1961)<br />Interviews contained hundreds of topics<br />I focus on a specific set<br />Began on the surface<br />Where in the context of the interview did people discuss the spread of HCV?<br />What did they say?<br />What is the range of responses?<br />What similarities were there?<br />
    20. 20. Quantitative Sample<br />Behavior Risk Factor Surveillance System (BRFSS), 2001, Arizona<br />Conducted by the Centers for Disease Control (CDC)<br />The world’s largest telephone survey<br />Nearly 200,000 people participated in 2001<br />Core questions<br />State questions. <br />Arizona is the only state to ask questions about what causes HCV. <br />
    21. 21. Measures in the CDC Survey<br />Sociodemographic Characteristics: Age, Sex, Race, Income and Education<br />Do you know someone diagnosed with hepatitis C?<br />Have you ever been tested for hepatitis C?<br />Do you consider yourself at risk of hepatitis C? <br />Have you ever been diagnosed with hepatitis C?<br />
    22. 22. Measures of the Knowledge of the Spread of HCV<br />Do you think hepatitis C can be spread thru?<br />Sneezing or Coughing<br />Kissing<br />Unprotected Sex<br />Food or Water<br />Sharing Needles to Inject Street Drugs<br />Using the Same Bathroom<br />Contact with the Blood of an Infected Person<br />
    23. 23. Methods of Quantitative Analysis<br />Managed with SPSS<br />Analyzed with MPlus<br />Analysis proceeded in several stages<br />Exploratory Factor Analysis <br />Confirmatory Factor Analysis<br />Latent Class Analysis (LCA)<br />Mixture Modeling (LCA with Covariates)<br />Corrected for measurement error<br />Robust estimation for binary indicators<br />Missing Values handled using multiple imputation.<br />
    24. 24. Qualitative Results<br />Where Patients Talked About HCV<br />How they themselves were infected with HCV<br />RJ: That part, I’m pretty sure I got it through drug use, intravenous. If, or, I should speak up too, drug use, you know.<br />How they think they could spread HCV to others <br />BT: I’m in a domestic violence shelter. I don’t use any of their dishes bowls, cups, I bought my own and I have paper, you know, little chinette plates and I use plastic forks and spoons because they’re disposable and I’m very, very careful because I don’t want to catch anything. I also don’t want to give anybody else anything. <br />How others think HCV can be spread <br />CG: About hepatitis C, guy tells me uh, “oh man don’t share my beer”. I says, “you know what, don’t share your beer”. I probably got it from him. (laughs) You know? You probably gave it to me. ‘<br />
    25. 25. Qualitative Results<br />What Patients Said<br />How they themselves were infected with HCV<br />RJ: That part, I’m pretty sure I got it through drug use, intravenous. If, or, I should speak up too, drug use, you know.<br />How they think they could spread HCV to others <br />BT: I’m in a domestic violence shelter. I don’t use any of their dishes bowls, cups, I bought my own and I have paper, you know, little chinette plates and I use plastic forks and spoons because they’re disposable and I’m very, very careful because I don’t want to catch anything. I also don’t want to give anybody else anything. <br />How others think HCV can be spread <br />CG: About hepatitis C, guy tells me uh, “oh man don’t share my beer”. I says, “you know what, don’t share your beer”. I probably got it from him. (laughs) You know? You probably gave it to me. ‘<br />
    26. 26. Qualitative Results<br /> “I don’t wear a sign saying, hey I got hepatitis C, you know, don’t use the bathroom. You know, even though I can’t transmit it that way.” Interview Respondent #28<br />[REVIEW CHAPTER FOR BETTER EXAMPLES]<br />
    27. 27. A Qualitative Example<br /> “One doctor told me they’ve discovered cases in people they figured they contracted back in the Korean War, you know. It’s, you don’t know you’ve got it. You know, it’s like high blood pressure. You know, people live for years with that. They don’t know they have it. Well the same thing with hepatitis C. It’s in there attacking your organs and you don’t have a clue it’s doing it. You don’t get jaundiced. You don’t really have any effects from it. So it’s not a constant reminder that I’ve got it. I don’t wear a sign saying, hey I got hepatitis C, you know, don’t use the bathroom. You know, even though I can’t transmit it that way.” Interview Respondent #28<br />
    28. 28. Distribution of Outcome Variables<br />
    29. 29. Description of Sample<br /><ul><li> Mean Age was 48.5 (SD 18.1)
    30. 30. 59% Female
    31. 31. 22% Know someone with HCV
    32. 32. 25% Said they had been tested for HCV
    33. 33. 75.8% White, 17.5% Hispanic, 4.4% Native American / American Indian, 2.3% Black
    34. 34. Mean Income Level was $50000-$75,000
    35. 35. Median Education Level was completion of high school</li></li></ul><li>Exploratory Factor Analysis<br />Unweighted Least Squares Estimation Varimax and Promax Rotation<br />Results with WLSMV Estimation were equivalent<br />Root Mean Square Residuals supported a two factor solution<br />
    36. 36. Exploratory Factor Analysis [CONSIDER SHOWING PROMAX]<br />EXPLORATORY ANALYSIS WITH <br />1 FACTOR :<br />ROOT MEAN SQUARE <br />RESIDUAL IS 0.0919<br />ESTIMATED FACTOR LOADINGS<br /> 1<br /> ________<br /> SEX2 0.800<br /> NEEDLE 0.890<br /> BLOOD 0.850<br /> SNEEZEC -0.743<br /> KISSC -0.892<br /> FOODC -0.808<br /> BATHROC -0.767<br />EXPLORATORY ANALYSIS WITH <br />2 FACTORS :<br />ROOT MEAN SQUARE <br />RESIDUAL IS 0.0222<br /> VARIMAX ROTATED LOADINGS<br /> 1 2<br /> ________ ________<br /> SEX2 0.760 0.357<br /> NEEDLE 0.929 0.339<br /> BLOOD 0.858 0.341<br /> SNEEZEC -0.212 -0.943<br /> KISSC -0.502 -0.771<br /> FOODC -0.516 -0.616<br /> BATHROC -0.456 -0.626<br />
    37. 37. CHECK DF FOR ACCURACY <br />That is probably a typo<br />
    38. 38. Fix spelling error above<br />
    39. 39. Three Latent Classes<br />2,4,5,6, and 7 category models do not fit the data as well as the 3 category model. <br />
    40. 40. Figure 4: Distribution of Forms of Knowledge (Most Likely Latent Class)<br />
    41. 41. Figure 5: Estimated Probabilities of Knowing How HCV is Spread by Most Likely Latent Class<br />
    42. 42. Structural Model Testing<br />What concepts are associated with the likelihood of each form of knowledge?<br />Mixture Modeling <br />Simultaneously test continuous and categorical predictors of class membership (form of knowledge). <br />Similar to Multinomial Logistic Regression<br />
    43. 43.
    44. 44. Results of Structural Model Testing<br />Strongest Associations<br />Race/ethnicity <br />Knowing someone with HCV<br />Feeling at risk for HCV <br />Not Associated<br />Gender<br />Having a health plan<br />Having had an HCV test<br />
    45. 45. Effects of Covariates on Form of Knowledge: Everywhere vs. True Awareness<br />N = 3092<br />
    46. 46. Effects of Covariates on Form of Knowledge: Nowhere vs. True Awareness<br />N = 3092<br />
    47. 47. Probability of Form of Knowledge by Race/Ethnicity (N = 3092)<br />
    48. 48. Probability of Form of Knowledge by Knowing Someone w/ HCV (N = 3092)<br />
    49. 49. Probability of Form of Knowledge by Education Level<br />
    50. 50. Probability of Form of Knowledge by Age<br />
    51. 51. Probability of Form of Knowledge by Whether R Had a Health Plan (N = 3092)<br />
    52. 52. Discussion<br />Most important excerpts from Ch. 7<br />Summary of surprising findings<br />
    53. 53. Limitations<br />Secondary Data<br />Limited Items about HCV <br />Arizona<br />Confidence in / certainty of knowledge<br />Sample Weights and Unequal Probability of Selection<br />Relationships among independent variables<br />
    54. 54. Conclusions<br />Descriptive<br />Racial and Ethnic Minorities are more likely to have the Everywhere and Nowhere forms of knowledge.<br />In the True Awareness group many respond that HCV is NOT transmitted sexually. <br />These individuals are not necessarily incorrect. Sexual transmission of HCV is infrequent overall, and virtually non-existent in studies of long-term monogamous couples. There is debate among scientists. <br />Methodological<br />Simple items assessing the causes and transmission of illness in a test-like format cannot always be assumed to have a continuous latent structure. <br />Practical<br />Lack of Awareness of HCV infection is likely to continue<br />Public health education<br />
    55. 55. Conclusions<br />Theoretical<br />Is the Nowhere form attributable to a simple ignorance of the scientific facts? Is it a desire not to know the facts? Is it the extremely high level of specialization in society? <br />The process of knowledge diffusion is linked to social and economic inequality.<br />Culture—in the form of knowledge of the spread of HCV—is connected to social structure. <br />
    56. 56. Final Thoughts<br />Ronald Reagan famously said of democrats in a 1964 speech<br />“Well, the trouble with our liberal friends is not that they&apos;re ignorant; it&apos;s just that they know so much that isn&apos;t so.”<br />In other words<br />Variation in lay people’s knowledge of the spread of HCV is culturally situated and cannot be viewed simply as part of a continuum of fidelity to biomedical and scientific evidence. <br />
    57. 57. References<br />Agency for Healthcare Research and Quality (AHRQ). 2003. Management of Chronic Hepatitis. Publication Number 02-E030. Available World Wide Web: http://www.ahrq.gov/clinic/epcsums/hepcsum.htm<br />Asparouhov, T. 2004. Weighting for unequal probability of selection in latent variable modeling. Mplus Web Notes: No. 7.<br />Bollen, K.A. 2002. Latent Variables in Psychology and the Social Sciences. Annual Review of Psychology, 53, pp. 605-634.<br />Centers for Disease Control (CDC) 2007. Hepatitis C Fact Sheet. Available WWW: http://www.cdc.gov/hepatitis <br />Gordon, F. D. 1999. Cost-effectiveness of screening patients for hepatitis C.&quot; The American journal of medicine 107.6B 36S-40S.<br />Hagenaars, J.A. 1998. Categorical Causal Modeling. Sociological Methods & Research, 26(4), pp. 436-486. <br />Kleinman, A. 1988. The Illness Narratives. New York: Basic Books. <br />Muthén, B. (2004). Latent variable analysis: Growth mixture modeling and related techniques for longitudinal data. In D. Kaplan (ed.), Handbook of quantitative methodology for the social sciences (pp. 345-368). Newbury Park, CA: Sage Publications.<br />Tataryn DJ, Wood JM and Gorsuch RL 1999. &quot;Setting the value of k in PROMAX: a Monte-Carlo study&quot;, Educational & Psychological Measurement, 59(3), pp.384-391.<br />
    58. 58. Adam T. Perzynski<br />Between Facts and Voices:Medical and Lay Knowledge of the Spread of Hepatitis C<br />E-mail: Adam.Perzynski@case.edu<br />This research was supported in part by NIH Grant No. 1 R01 AA13302-01A1<br />

    ×