Health information behavior in families: supportive or irritating?

1. Health information behavior in families: supportive or irritating? Tiffany Veinot Yong-Mi Kim Chrysta Meadowbrooke

2. INTRODUCtion

3. Chronic disease • Chronic diseases: prolonged illnesses that can be controlled but not cured • Family members are important resources for chronically ill people • Social support • Coping assistance (Thoits, 1995) • Informal care (Barrett, 2004) • Information

4. Social support • Social support: • “social resources” that a person perceives as available or that he or she actually receives from non-professionals in their lives • Perceived support • “…beliefs about the availability of varied types of support from network associates” • Enacted support • “reports about the types of support received” (Gottlieb & Bergen, 2010).

5. Social support • For people with diabetes or HIV/AIDS, social support has been associated with: • better mental health (e.g., Harkness et al., 2010; Reich, Lounsbury, Zaid-Muhammad, & Rapkin, 2010) • improved health outcomes (disease progression, mortality) (e.g., Zhang, Norris, Gregg, & Beckles, 2007; Leserman et al., 2000)

6. Informational support • “... giving information, advice, suggestions, or directives.” (Uchino, 2004) • But disease-related knowledge of lay people may vary by illness. • Prevalence within groups (Matsagianis, 2008; Warren-Findlow & Prohaska, 2008). • Stigmatization (Veinot, 2010; Veinot & Harris, 2011).

7. Hypotheses 1-2 • H1: People with diabetes will have significantly larger family-based networks for enacted social support than people with HIV/AIDS (PHAs), including larger family networks for informational support. • H2: People with diabetes will report significantly more perceived social support (PSS) from their family members than PHAs.

8. Informational & emotional support • Typically conceptualized as distinct (Harlow & Cantor, 1995). • But often occur together and can be difficult to separate (Berkman & Glass, 2000). • Strong association between these two forms of support in close relationships (Dunkel-Schetter, Folkman, & Lazarus, 1987; Schaefer, Coyne, & Lazarus, 1981). • Informational support may “signal caring” (Schaefer, 1981).

9. Hypothesis 3 • H3: Among both diabetics and PHAs, enacted informational support from family members will be significantly and positively associated with their receipt of emotional support from these relatives.

10. Collaborative information behavior • Information support as “information sharing”. • Collaborative information behavior (CIB): • “an activity where two or more actors communicate to identify information for accomplishing a task or solving a problem” (Talja & Hansen, 2006, p. 114). • Also includes looking for new information. • May be seen as supportive (Veinot, 2009).

11. Hypotheses 4-5 • H4: People with diabetes will report significantly more CIB with their family members than PHAs. • H5: Among both diabetics and PHAs, CIB will be positively and significantly associated with PSS.

12. Information use • Information use has not been a focus in previous social support research. • Usefulness may vary by family members’ knowledge and experience. • Information that is used may be seen as more supportive than information that is not used.

13. Hypotheses 6-7 • H6: People with diabetes will report significantly more use of information obtained collaboratively with their family members than PHAs. • H7: Among both diabetics and PHAs, use of information obtained collaboratively will be positively and significantly related to PSS.

14. Research question • RQ1. Why do study variables have these statistically significant relationships?

15. METHODS

16. The Study • Preliminary results of a longitudinal, mixed-methods study. • Results reported here for T1 data collection with first 24 patient recruits. • Quantitative surveys. • Concurrent qualitative interviews.

17. Participants • 12 diabetics and 12 people with HIV/AIDS . • Recruited from health clinics and community-based, non-profit agencies. • 70 people approached at clinics; 24% response rate (RR). • Non-profit recruitment emphasized gender and racial diversity (no RR available).

18. Procedures • Three part meeting lasting 1-1.5 hours. • Paper survey. • Demographics, PSS, and information behavior. • In-depth, semi-structured interview. • Family experience of managing HIV/AIDS or diabetes. • Visualization of family support networks (Marin & Hampton, 2007).

19. Variables – Social support • Perceived social support (PSS). • “Family” factor of Multidimensional Scale of Perceived Social Support (MSPSS) (Zimet, Dahlem, Zimet, & Farley, 1988). • Cronbach’s α = .92 • Enacted social support. • Emotional support network size + Tangible support network size + Informational support network size.

20. Variables – Information behavior • Collaborative information behavior (CIB) frequency. • New 13-item scale. • E.g., “I looked on the Internet with a family member”. • Cronbach’s α = .90. • Information use frequency. • New 16-item scale. • E.g., “Feel more in control of my diabetes”. • Cronbach’s α = .94.

21. Data analysis – survey • Descriptive statistics. • Tests of differences between PHAs and diabetics: t-tests, a Mann-Whitney U test. • Bivariate associations between variables: • Pearson product-moment correlations. • Chi-square tests.

22. Data analysis – interviews • Interviews were audio-taped and transcribed. • We created participant profiles and “open coded” transcripts (Strauss & Corbin, 1998): • Choices affecting support network size. • Statements re: emotional-informational support relationship. • Family-based CIB. • Theme frequencies by survey responses.

23. Results

24. Demographics (n=24)

25. Demographics (n=24)

26. Demographics • Mental health and/or substance abuse. • 10 PHAs and 3 diabetics • Diagnosis was dependent on mental health or substance abuse history (χ2 (1, N=24) = 6.00, p < .05, phi = .50). • Self-rated physical health: good to excellent. • 10 diabetics and 12 PHAs • Diagnosis was independent of self-related health (χ2 (3, N=24) = 4.22, p = .239, phi = .42)

27. Demographics • At least one other person in their family with their illness. • 8 diabetics and 2 PHAs. • Diagnosis was dependent on having a person in one’s family with one’s illness (χ2 (1, N=24) = 8.71, p < .01, phi = .60).

28. Family network size by disease • The network size difference was statistically significant (t(22)=2.43, p=.024). • The information network size difference was marginally significant (t(22)=1.85, p=.078). • Hypothesis 1 was partially supported.

29. PSS by disease • People with diabetes had higher mean PSS scores than PHAs (M = 4.25 vs. 3.97). • This difference was not statistically significant (t(22)=.97, p=.34). • Hypothesis 2 was not supported.

30. Explaining the difference • People with diabetes had more family members with their disease than PHAs. • Family frequently shared emotional and informational support with one another. • “…the loss of sight in my right eye…that’s because of the diabetes… we were talking because she’s having some eye problems now…we were…looking up different things …so that hopefully she won’t develop the same thing.”

31. Explaining the difference • More diabetics than PHAs had ties with immediate family (U = 30.50, z = -2.43, p < 0.05) or “other” (U = 33.00, z = -2.33, p < 0.05).

32. Explaining the difference • Mental health and substance abuse history. • Three PHAs had consciously limited network size due to mental health or addictions. • Only one diabetic had done so.

33. Explaining the difference • More PHAs than diabetics reported significant interpersonal constraints in speaking with their family members about their disease (9/12 PHAs vs. 1/12 diabetics). • Decision not to disclose illness to family. • “…none of my blood family knows. I…prefer to keep it that way…because when I came out to my mom, one of the big things that she was concerned about was that I would get HIV…”

34. Explaining the difference • Negative family response to illness. • Three PHAs were estranged from judgemental relatives. • “…I have been, like, the punching bag and the point of jokes… in their home, I was the easy target to lambast…” • Two PHAs had family members who knew, but never spoke of it. • “…he knows I have HIV. He never asked me about it …I think he’s disappointed…”.

35. Information & emotional support • A person who provides informational support was more likely to provide emotional support (or vice versa) • This was true for both people with diabetes, (χ2 (1, N=117) = 25.42, p < .001, phi = .48), and PHAs, χ2 (1, N=66) = 12.20, p < .001, phi = .47. • Hypothesis 3 was supported.

36. Explaining the relationship • Questions of concern. • Goal attainment assistance. • “…it’s not only helpful because of the information, but it’s helpful because of the support that I’m getting from them. It’s like stroking my ego…” (laughs) • Caring e-mail forwards. • “…give me hope…I know that’s what they’re trying to do.”

37. Collaborative information behavior • No significant differences between PHAs’ and diabetics’ collaborative information behavior (CIB) frequency scores (t(22)=.43, p=.67). • Hypothesis 4 was unsupported.

38. Collaborative information behavior • For diabetics, there was a negative correlation between CIB and PSS (r=-.59, N=12, p<.05). • For PHAs, this relationship was positive (r=.58, N=12, p=.05). • Hypothesis 5 was unsupported. • But this surprising difference warranted further investigation.

39. Explaining the relationship • Negative information sharing experiences. • Differences in attempted influence (8 diabetics vs. 0 PHAs), esp. re: food. • “…first thing she screams is, ‘How’s your diabetes? Have you taken your medication?’… every once in a while it’s a little too much caring, and it can get really irritating.”

40. Explaining the relationship • Often felt criticized or infantilized by family attempts to influence them. • “…[my wife] holler[s] at me if I pick up a donut … I don’t like to be told what to do… if she tells me ‘no,’ I’ll eat two of them.” • Unwanted interference in relationships with others (e.g., doctors). • “She brings up things about my health care that concern her that don’t concern me…at times it feels interfering.”

41. Information use • No significant differences between PHAs’ and diabetics’ information use frequency scores (t(22)=.59, p=.56). • Hypothesis 6 was unsupported.

42. Information use • For diabetics, there was no significant relationship between information use and PSS (r=-.23, N=12, p=.47). • There was also no significant relationship for PHAs (r=.54, N=12, p=.07). • Hypothesis 7 was unsupported.

43. DISCUSSION

44. Discussion • This study among the first to show differences in social support and CIB among people with different diseases. • Further research needed with larger sample. • Provision of information systems and services for both ill people and their families.

45. Discussion • Differentiation of information systems / services by disease. • For less prevalent, more stigmatized diseases like HIV/AIDS: • More outreach, promoting dialogue and improved attitudes, expanding social support networks • For more prevalent, less stigmatized diseases like diabetes: • Information literacy and management, resolving disagreements, etc.

46. Discussion • Overlap of information and emotional support in close relationships. • Information and emotional support may be the same thing in certain types of interactions. • Thus even information of poor quality, or irrelevant information, was often appreciated.

47. Discussion • Relational significance of information sharing and information systems / services. • Organizing, presenting or rating content according to emotional valence. • Referral to both factually and relationally helpful information.

48. Discussion • Negative relationship between CIB and PSS finds resonance in health psychology. • Misunderstanding or social control. • Information is not necessarily supportive or helpful. • Previously defined as a “help” or “resource” (Dervin, 1992; Harris & Dewdney, 1994; & Rice, 1999). • Coaching or training in information provision?

49. Limitations • Small, non-randomly selected sample of only 24 participants. • Insufficient statistical power for multivariate statistical tests. • Cross-sectional data.

50. conclusion