Public Preferences about Secondary Uses of Electronic Health Information

Abstract

Importance

As health information technology grows secondary uses of personal health information offer promise in advancing research, public health, and health care. Public perceptions about personal health data sharing are important to establish and evaluate ethical and regulatory structures for overseeing the use of these data.

Objective

Measure patient preferences toward sharing their electronic health information for secondary purposes—uses other than their own health care..

Design

In this conjoint analysis study, participants were randomized to receive 6 of 18 scenarios describing secondary uses of electronic health information, constructed with 3 attributes: uses (research, health care quality improvement, marketing), users (university hospital, drug company, public health department), and data sensitivity (medical history, medical history plus genetic test results). This experimental design enabled participants to reveal their preferences for secondary uses of their personal health information.

Setting and Participants

We surveyed 3,336 Hispanic (n=568), non-Hispanic African American (n=500), and non-Hispanic White (n=2,268) adults representing 65.1% of those from a nationally representative, online panel.

Main Outcomes and Measures

Participants responded to each conjoint scenario by rating their willingness to share their electronic personal health information on a 1–10 scale (1=low, 10=high). Conjoint analysis yields importance weights reflecting the contribution of a dimension (use, user, sensitivity) to willingness to share personal health information.

Results

The use of data was the most important factor in the conjoint analysis (63.4% importance weight) compared to the user (32.6% importance weight) and data sensitivity (importance weight: 3.1%). In unadjusted models, marketing uses (−1.55, p<0.001), quality improvement uses (−0.51, p<0.001), drug company users (−0.80, p<0.001) and public health department users (−0.52, p<0.001) were associated with less willingness to share health information compared to research (use) and university hospitals (users). Hispanics and African-Americans discriminated less between the three uses compared to Whites.

Conclusions and Relevance

Participants cared most about the specific purpose for using their health information, though differences were smaller among racial and ethnic minorities. The user of the information was of secondary importance and the sensitivity of information was not a significant factor. These preferences should be considered in policies governing secondary uses of health information.

Introduction

Over the past four years, the federal government has made an unprecedented public investment in health information technology (HIT). By the end of 2012, 72% of office-based physicians had adopted an electronic health record system.¹ HIT policy discussions have generally focused on how HIT adoption can affect the quality and value of health care.² However, increasingly digitized health information also enables new and potentially far-reaching opportunities for secondary uses of electronic health information. Secondary uses, which we define as uses other than personal medical care, fall into several categories including: research (e.g. comparative effectiveness studies), public health (e.g. public health surveillance), health care quality improvement (e.g. measurement of provider performance), and commercial marketing (e.g. pharmaceutical marketing).

There are strong ethical traditions (i.e. autonomy) and regulatory structures (i.e. privacy laws) that emphasize the privacy of this information and the rights of patients to know about and approve its use.³ However, there are circumstances when privacy is overridden to advance societal interests such as the case of certain contagious or otherwise reportable illnesses.⁴ Ethicists and regulators have debated other circumstances where the proper balance between individual privacy interests and public good is less clear. For example, comparative effectiveness research could greatly advance scientific understanding of health and health care if it could be performed on clinically detailed and broadly representative information from interoperable electronic health records. The social benefits of such research might be large enough to justify more relaxed approaches to consent if inviolable elements of personal privacy were preserved.^4–7 De-identifying data does not necessarily eliminate these tensions between personal privacy and public good – first, because the highly protective systems of de-identification imposed under HIPAA often eliminate information important to retain^8,9 and second, because patients often express concern about data stewardship and control separate from issues of privacy.^10,11

In this context, we sought to investigate public attitudes about the secondary use of electronically available personal health information. We recognized that these attitudes themselves will differ according to context and so we deployed conjoint analysis in an experimental survey design to examine patients’ willingness to share their personal health information. Conjoint analysis is commonly used in marketing research to disentangle preferences for individual attributes of consumer products when presented in combination,^12–14 for example, automobiles that vary according to price, performance, comfort, gas mileage, safety, and style. By asking individuals to rate different combinations of these attributes, it is possible to measure the unique importance of each attribute to the consumer. In the current context, the attributes are not consumer product characteristics but attributes of how health information might be used. We examined different uses of this information (medical research, quality improvement, commercial marketing), different users of this information (university hospitals, public health departments, commercial enterprises), and different personal sensitivity of this information (whether it included genetic information about their own cancer risk). Prior studies^10,15–22 suggest that each of these factors are potentially important determinants of public attitudes about data use but no studies have examined them systematically. We focused on genetic information and cancer risk given the heightened sensitivity to this type of health information and the strong interest in building genomic databases for future research.^23,24 We also focused on three racial and ethnic subpopulations, non-Hispanic Whites, non-Hispanic African-Americans and Hispanics, using responses as a lens to examine differences across these broad groups given the historical relationship between race, genetics, and distrust in biomedical research.^25–28

Methods

Participants

We recruited participants from an online research panel assembled by GfK Knowledge Networks.²⁹ The GfK Knowledge Networks panel members are recruited through probability-based sampling using both random digit dialing and address-based sampling to create nationally representative panels.³⁰ This method provides coverage of households with and without telephones, cellphones, and Internet access (97% coverage of U.S. households) Individuals without Internet access are provided computer hardware and Internet access to complete surveys. Individuals with Internet access are provided modest cash compensation. On average, panel members are invited to participate in 4–6 surveys per month and complete 3 surveys per month. Online research panels constructed from probability-based samples have been shown to yield similar estimates to random digit dial telephone surveys.³¹

We over-sampled African-Americans and Hispanics (from English and Spanish speaking households) to allow for comparisons across racial and ethnic groups. We administered the survey electronically from November 9, 2012 to December 2, 2012. Prior to analysis we excluded respondents who completed the survey in less than half the median completion time (<5 minutes) to screen out “speeders” who were very unlikely to have read the questions prior to responding and who, in pretest analyses in this population revealed almost no item to item variation in responses consistent with inattention to question content. The protocol was approved by the University of Pennsylvania Institutional Review Board.

Experimental Instrument

The online survey instrument contained 3 main sections: [1] health status and health care access measures; [2] responses to short scenarios describing secondary data uses; and [3] a 9-item scale measuring health care system distrust.

Conjoint Scenarios

Participants were presented with the following introduction:

Many doctors and hospitals are starting to use electronic medical records instead of paper charts when they provide care. Electronic medical records can also be used for other health care and public health reasons.

You will be shown some possible uses of your electronic health information. In each case, you will be shown what health information will be used, who will use it, and what they will use it for. Please indicate how willing you would be to share your health information for each situation. Your name would not be released.

This introduction was followed by descriptions of 6 different situations in which participants were told what information would be used reflecting the sensitivity of the information (“your medical history” (i.e., low sensitivity) or “your medical history and the results of a personal genetic test that predicts your chance of getting cancer” (i.e., high sensitivity)), the user of the information (“a university hospital,” “drug company,” or “public health department”) and the use of the information (“research new ways to prevent cancer,” “evaluate how well your doctor provides preventive cancer care,” or “identify what kinds of patients will be interested in buying their cancer prevention product”) Each participant was randomly assigned to receive six out of the possible eighteen scenarios. For each item, participants were asked “In this situation, how willing would you be to share your health information?” rated on a 10-point scale (1=not at all, 10=very willing).

Health and Health Care Measures

Prior research suggests health status may influence support for sharing health information; however, the evidence is mixed.^15,19 We measured self-rated health with a single item from the Behavioral Risk Factor Surveillance System (BRFSS) 2010 questionnaire (SF-1).³² We evaluated insurance status, whether the respondent had a usual source of medical care, and whether the respondent had experienced cost-related barriers to care in the prior year using items derived from the National Health Interview Survey.³³ We hypothesized that individuals marginalized from the health care system or those with a strained financial relationship would be less willing to share their health information. We measured distrust in the health care system using the Revised Health Care System Distrust Scale³⁴ – a 9-item scale ranging from 9–45 (45=high distrust) including two validated subscales: competence distrust and values distrust. Concerns about health information privacy potentially relate to both domains if patients are concerned about mishandling of information as well as uses of information that may bring harm through discrimination or stigma.^35,36 Distrust may also mediate associations between race/ethnicity and willingness to share health information given the history of distrust among minority populations about biomedical research.^25–28

Confidence in Institutions and Organizations

We asked respondents to rate various institutions and organizations on how much confidence they had in each to protect their health information. Respondents chose from three categories: hardly any confidence at all, only some confidence, very high confidence. This question was adapted from a General Social Survey question measuring confidence in institutions overall.³⁷

Demographics

GfK Knowledge Networks had previously collected demographic data on panel participants including age, sex, race, ethnicity, income, education, and metropolitan-rural residential status.

Statistical Analysis

We compared characteristics of responders and non-responders and compared the sociodemographic characteristics between racial/ethnic groups (non-Hispanic White, non-Hispanic African American, and Hispanic) using t tests and ANOVA for continuous variables and chi-squared tests for categorical variables.

We conducted a conjoint analysis based on a main effects analysis-of-variance model. In this analysis, we computed a numerical part-worth utility value for each level of each attribute.^13,38 Large part-worth utilities were assigned to the most preferred levels, and small part-worth utilities were assigned to the least preferred levels. The attributes with the largest part-worth utility range were considered the most important in explaining the variability in the outcome. Importance weights are reported for each attribute.

Subsequently, in order to adjust for patient level characteristics and to investigate potentially important interactions with the attributes, we conducted linear regression between the attribute levels, participant characteristics and distrust. Correlation between the conjoint scenarios nested within respondent was accounted for using generalized estimating equations assuming an independence correlation structure. Post-stratification weights provided by GfK Knowledge Networks were used in all analyses to adjust for any survey non-response as well as any non-coverage or under- and over-sampling resulting from the study-specific sample design.

Models include the three test attributes, patient race/ethnicity, a measure of high health care distrust (top quartile), health status, measures of access to care, and other socio-demographic variables, as well as interaction terms between the attributes and patient race/ethnicity. We specified a baseline scenario in the models where we hypothesized support would be the highest (sensitivity: low; user: university hospital; use: research). Wald tests were used to assess significance with a type I error rate of 0.05. All analyses were conducted in SAS 9.3 (SAS Institute Inc., Cary, NC, USA).

Results

Of the 5,119 panel members we invited to participate, 3,336 completed the survey (response rate: 65.1% - AAPOR method 1) and 272 “speeders” were removed for a final sample of 3,064 which included 2,093 non-Hispanic Whites, 500 non-Hispanic African Americans, and 516 Hispanics (250 completed the survey in Spanish). The socio-demographic, health, and health care variables differed across the three racial and ethnic categories (Table 1). There were differences in the demographic characteristics of responders and non-responders. Non-responders were younger, lower income, lower education, and slightly more likely to be female. However, we applied post-stratification weights to our analyses that account for non-response bias.

Table 1.

Participant Characteristics (unweighted percentages)

	Overall (n=3,064)	African American (n=455)	Hispanic (n=516)	White (n=2,093)	P- Valuea
Age					<0.001
18–29	13.2	13.9	22.5	10.7
30–44	21.5	20.7	32.0	19.1
45–59	32.4	32.5	28.9	33.3
60+	32.9	33.0	16.7	36.9
Female (%)	49.8	53.6	49.0	49.2	<0.001
Lives in a metropolitan area (MSA)	85.7	91.7	94.0	82.4	<0.001
Education					<0.001
Less than high school	8.9	9.5	24.0	5.1
High school	29.1	29.7	29.7	28.9
Some college	30.4	36.7	27.7	29.6
Bachelor's degree or higher	31.6	24.2	18.6	36.5
Income					<0.001
<$25,000	18.8	29.9	25.8	14.7
$25,000–49,999	23.6	25.3	31.2	21.3
$50,000–74,999	19.0	17.6	18.6	19.5
$75,000 and above	38.6	27.3	24.4	44.5
Insured	82.9	80.6	59.1	89.3	<0.001
Has a personal doctor or other provider	75.8	77.8	56.2	80.2	<0.001
Did not receive care in past year due to cost	16.8	21.7	26.1	13.4	<0.001
Health Status: Fair/Poor	16.4	21.0	20.4	14.4	<0.001

^aANOVA results comparing results across the three racial/ethnic groups.

The mean willingness to share health information for each conjoint scenario is presented in Table 2 by race/ethnicity. In our baseline scenario where we hypothesized support would be the highest (sensitivity: low; user: university hospital; use: research), overall willingness to share electronic health information was moderately high (mean: 6.82; Scale: 1–10, 1=not at all willing to share, 10=very willing to share). For this baseline scenario, Whites demonstrated greater willingness to share their electronic health information compared to Hispanics (mean: 6.98 vs. 6.38, p=0.027). The differences between Whites and African Americans were not significant (mean: 6.98 vs. 6.58, p=0.21).

Table 2.

Respondent Willingness to Share Personal Health Information by Conjoint Scenario (unadjusted means)

Conjoint Scenario			Willingness to Share Personal Health Information (1=low, 10=high)
Sensitivity	Use	User	Overall	Non- Hispanic African- American	Hispanic	Non- Hispanic White	P- Values*
Low	Research	University Hospital	6.82	6.58	6.38	6.98	0.04
High	Research	University Hospital	6.72	6.60	6.80	6.73	0.84
Low	Research	Drug Company	5.90	5.71	5.92	5.94	0.84
High	Research	Drug Company	5.86	5.79	5.79	5.90	0.86
Low	Research	Public Health Dept.	6.10	6.23	6.14	6.06	0.84
High	Research	Public Health Dept.	6.18	6.07	6.12	6.22	0.84
Low	Quality Improvement	University Hospital	6.33	6.64	6.55	6.20	0.16
High	Quality Improvement	University Hospital	6.32	6.11	6.53	6.31	0.57
Low	Quality Improvement	Drug Company	5.06	5.36	5.51	4.90	0.04
High	Quality Improvement	Drug Company	5.33	5.63	5.75	5.15	0.05
Low	Quality Improvement	Public Health Dept.	5.85	5.72	6.17	5.80	0.33
High	Quality Improvement	Public Health Dept.	5.64	6.18	5.99	5.47	0.03
Low	Marketing	University Hospital	4.87	5.29	5.55	4.62	<0.001
High	Marketing	University Hospital	5.05	5.43	5.32	4.90	0.08
Low	Marketing	Drug Company	4.53	5.45	5.01	4.23	<0.001
High	Marketing	Drug Company	4.63	5.24	5.30	4.36	<0.001
Low	Marketing	Public Health Dept.	4.53	4.73	5.46	4.24	<0.001
High	Marketing	Public Health Dept.	4.67	5.14	5.43	4.40	<0.001

^*p-values reflect comparisons across the three racial/ethnic categories using Benjamini-Hochberg multiple comparisons correction method

Using conjoint analysis, we calculated importance weights for the three attributes. Importance weights are a measure of the importance of an attribute relative to the other attributes in the model on an individual’s preferences. The use of data was the factor that most influenced participants’ willingness to share their electronic health information (64.3% importance weight). The user of the data was less important (32.6% importance weight). The sensitivity of health information was not important (3.1% importance weight). Use was most important among all three racial/ethnic groups (importance weights – White: 65.7%, Hispanic: 57.9%, African-American: 56.4%) compared to user (importance weights – White: 31.5%; Hispanic: 38.3%; African American: 41.3%) and sensitivity (importance weights – White: 2.7%; Hispanic: 3.8%; African American: 2.3%).

In unadjusted GEE models, use and user variables were statistically significant as shown in Table 3. Marketing uses resulted in the largest declines in willingness to share health information compared to research uses (β: −1.55, p<0.001). Respondents also reported less support for quality improvement uses compared to research uses (β: −0.51, p<0.001). The user of health information was also important. Drug companies (β: −0.80, p<0.001) and public health departments (β: −0.52, p<0.001) received lower support compared to university hospitals. We tested interactions of race/ethnicity with the conjoint attributes (use, user, sensitivity) while controlling for socio-demographic characteristics including education and income, health status, and access to care measures. The difference in willingness to share health information for marketing versus research uses was smaller for African Americans and Hispanics compared to Whites (African American*marketing interaction (β): +0.87, p<0.001; Hispanic*marketing interaction (β): +1.01, p<0.001). There was also a smaller but statistically significant interaction between race/ethnicity with quality improvement uses compared to research uses (African American*quality improvement interaction (β): +0.44, p=0.001; Hispanic*quality improvement interaction (β): +0.59, p<0.001). In Figure 1, we display these interactions for the three uses (research, quality improvement, and marketing) for two users (university hospital and drug company) by race/ethnicity.

Table 3.

Influence of attributes of health information sharing on willingness to share health information

	Model 1a			Model 2a			Model 3a
	β	P	95% CI	β	P	95% CI	β	P	95% CI
Conjoint Attributes
Use: Research	ref
Use: Quality Improvement	−0.51	<0.001	[−0.60, −0.42]	−0.67	<0.001	[−0.78, −0.56]	−0.68	<0.001	[−0.78, −0.57]
Use: Marketing	−1.55	<0.001	[−1.65, −1.44]	−1.84	<0.001	[−0.97, −1.72]	−1.85	<0.001	[−1.98, −1.73]
User: University hospital	ref
User: Drug company	−0.80	<0.001	[−0.90, −0.71]	−0.88	<0.001	[−1.00, −0.77]	−0.88	<0.001	[−1.00 −0.76]
User: Public health department	−0.52	<0.001	[−0.62, −0.43]	−0.60	<0.001	[−0.71, −0.48]	−0.59	<0.001	[−0.71, −0.48]
Sensitivity: Med history	ref
Sensitivity: Med history + genetic results	0.05	0.21	[−0.03, 0.12]	0.05	0.20	[−0.03, 0.14]	0.09	0.08	[−0.01, 0.20]
Race-Ethnicity
White				ref
African-American				−0.28	0.12	[−0.62, 0.07]	−0.35	0.05	[−0.70, 0.00]
Hispanic				−0.28	0.11	[−0.62, 0.06]	−0.30	0.10	[−0.66, 0.06]
Interactions (Race/Ethn*Attributes)
Drug Company * African American				0.28	0.05	[0.00, 0.56]	0.26	0.07	[−0.02, 0.54]
Drug Company * Hispanic				0.23	0.10	[−0.04, 0.51]	0.22	0.12	[−0.06, 0.50]
Public Health Dept * African American				0.15	0.27	[−0.11, 0.42]	0.15	0.28	[−0.12, 0.42]
Public Health Dept * Hispanic				0.29	0.025	[0.04, 0.54]	0.26	0.05	[0.00, 0.52]
Quality Improvement * African American				0.44	<0.001	[0.18, 0.70]	0.44	0.001	[0.18, 0.70]
Quality Improvement * Hispanic				0.56	<0.001	[0.32, 0.81]	0.59	<0.001	[0.34, 0.84]
Marketing * African American				0.88	<0.001	[0.58, 1.19]	0.87	<0.001	[0.57, 1.18]
Marketing * Hispanic				1.00	<0.001	[0.72, 1.29]	1.01	<0.001	[0.72, 1.30]
Med History + Genetic Test * African Am				−0.01	0.89	[−0.21, 0.19]	−0.01	0.88	[−0.21, 0.19]
Med History + Genetic Test * Hispanic				−0.01	0.9	[−0.22, 0.19]	−0.02	0.87	[−0.22, 0.19]
Covariates
Age
18–29							ref
30–44							−0.25	0.08	[−0.54, 0.03]
45–59							−0.38	0.008	[−0.65, −0.10]
60+							−0.23	0.12	[−0.52, 0.06]
Female							0.11	0.21	[−0.06, 0.29]
Lives in a metropolitan area (MSA)							0.16	0.20	[−0.09, 0.42]
Education
Less than high school							ref
High school							0.01	0.94	[−0.33, 0.36]
Some college							−0.03	0.85	[−0.39, 0.32]
Bachelor's degree or higher							−0.09	0.63	[−0.46, 0.28]
Income
<$25,000							ref
$25,000–49,999							0.09	0.53	[−0.19, 0.37]
$50,000–74,999							−0.13	0.39	[−0.43, 0.17]
$75,000 and above							0.02	0.91	[−0.27, 0.30]
Uninsured							0.05	0.64	[−0.24, 0.34]
Does not have a personal doctor or provider							−0.44	<0.001	[−0.67, −0.20]
Did not received care in past year due to cost							−0.30	0.024	[−0.56, −0.04]
Health Status: excellent, very good, or good							−0.11	0.51	[−0.42, 0.21]

^aLinear regression models using generalized estimating equations to account for correlation between the conjoint scenarios nested within respondent. The outcome is “willingness to share health information” on a 1–10 scale (1=low, 10=high).

Figure 1. Effect of Health Information Sharing Attributes on Willingness to Share by Race.

The Figure shows six of the eighteen conjoint scenarios. In each, the sensitivity of the health information is held constant (“low”). The user, “public health department,” is not shown here. The Figure shows the interaction of race and ethnicity with the conjoint attributes. The coefficients represent changes in willingness to share health information on a 1–10 scale (1=low, 10=high).

Respondents lacking a usual source of care (β: −0.44, p<0.001) and those experiencing cost barriers to care (β: −0.30, p=0.024) were less willing to share their health information. Participants reporting high levels of health care system distrust were less willing to share their health information (β: −0.65, p<0.001); however, distrust did not influence the relationship between race and the conjoint attributes. There were no significant differences by demographic characteristics except that young adults were slightly more willing to share their health information compared to middle aged adults.

We asked respondents how confident they were in various institutions and organizations to protect their health information. The results varied widely and are presented in Figure 2. Health care provider organizations and non-profit health-related organizations received relatively high support while internet and social media companies had very low support.

Figure 2. Confidence in Various Institutions and Organizations to Protect Health Information.

Respondents were asked, “Next, we are going to name some institutions and organizations in the country. How much confidence do you have in them to protect your health information?” They were offered three response options which are shown in the Figure.

Comment

As clinically detailed elements of personal health information are increasingly coded in electronic health records and as those records become more aggregatable across systems, secondary uses of electronic health information offer enormous promise in advancing research, public health, and health care. Even if we see the use of this information as providing immense social good, its effective use will likely require the acceptance of the public, and trust and confidence about information stewardship. This study begins to describe population support for specific uses of personal health information. Our study is based largely on revealed preferences derived from an experimental design, rather than expressed preferences, and has three main findings.

First, what our participants care about most is the specific purpose for using information, and among the choices we investigated, the goal they most privilege is research. Although it is not surprising that support was greater for research uses than marketing uses, it is surprising that support for quality improvement uses was lower than research since it is the use we described that is most likely to offer a more immediate benefit to the patient. Participants also revealed they care about the user, but less than use. To our surprise, the sensitivity of the health information—at least within the range we presented (personal medical history + personal genetic test results versus personal medical history alone) --was not important. This finding contrasts with the notion that patients view genetic information as particularly sensitive.^39,40 This finding may add support to the arguments against privileging genetic information as some experts have argued.^41,42

Second, we found that compared to White participants, racial and ethnic minorities were no less supportive of their health information being used for any use except for a small but non-significant difference for research. We had hypothesized that racial and ethnic minority respondents would be less supportive given prior work showing lower levels of support for research participation and higher levels of concern for HIT.^{15,25–28,43}

Third, relationships with the health care system are associated with support for sharing health information. Participants with high levels of health care system distrust, those without a usual source of care, and those that had recently experienced cost barriers to care were less supportive of secondary uses. However, participants also identified the health care delivery system as the most trusted to protect health information relative to other institutions and organizations—although less than half of participants expressed a high level of confidence even in these organizations.

This study has several limitations. First, we presented participants with hypothetical scenarios which may not reflect decisions in more naturalized circumstances. However, responses to hypothetical scenarios have a high level of concordance with actual behavior.⁴⁴ Second, we were not able to test an exhaustive list of attributes describing different data use scenarios. For example, it is possible that we might have found differences by sensitivity of health information if we explored a broader range of examples with potentially greater (e.g. diagnosis of HIV) or lesser (e.g. height) sensitivity. It is also possible that respondents think differently about the sensitivity and risks associated with genetic information compared to other health information thought to be sensitive. Third, our respondents may not be fully representative of the US population of the three racial and ethnic groups we sampled and there may be important differences within those broad racial/ethnic categories that this study cannot uncover. In addition, there may be differences in the attitudes of those willing to participate in a survey research panel compared to those who are not.

This study also has strengths. In particular, we used conjoint analysis in a carefully controlled experimental design enabling participants to reveal their preferences rather than merely express their preferences in response to static questions which might otherwise be subject to attribution bias or social desirability bias. In contrast, revealed responses provide a more naturalized context, even when presented in hypothetical settings. Indeed, the experimental design of our survey gives us greater confidence that the effects we observe within the sample are due to true effects of the conjoint attributes rather than other unobserved factors.

Most policy discussions about secondary uses of health information focus on whether data is identifiable or not. If it is de-identified, much of its value can be lost. If it is identifiable, restrictive privacy regulations sometimes limit the potential applications of those data toward social good. Although discussions about data identification are important, our study suggests that populations care about the purpose for data use and so these considerations might be elevated in decisions about personal health data sharing. As the use of health information technology expands, so will the potential value of the health information generated for uses beyond personal medical care. In addition, as more health information is generated outside of traditional medical encounters (e.g. patient health records), so too will the pressure to define acceptable uses. Our research can help inform these future debates as well as those focused on human subjects protections such as proposed changes to the Common Rule governing human subjects protections.⁴⁵ Our study also suggests that organizations and institutions within the health care delivery system are the most trusted stewards of health information and should play an instrumental role in efforts to extend the uses of population-level health information.