Abstract
This article explores the effect of doctor–patient communication about cancer-related Internet information on self-reported outcomes. Two hundred and thirty cancer patients and caregivers completed an online survey regarding their experiences searching for and discussing with their doctors cancer-related Internet information. Participants who assertively introduced the Internet information in a consultation were more likely to have their doctor agree with the information. When doctors showed interest and involvement and took the information seriously, participants were less likely to report a desire to change the doctor's response. Taking the information seriously was also associated with greater satisfaction. This preliminary evidence that the doctor's response is associated with patient outcomes indicates the potential for improving patient-centered communication. In an effort to maximize patient-centered communication, doctors should be encouraged to take their patients and the information they present seriously, as well as show their patients that they are interested and involved.
Keywords: oncology, provider-patient communication, Internet, patient satisfaction, patient-centered communication
INTRODUCTION
Eighty percent of adult Internet users seek health information from the Internet (Fox, 2006). Consequently, given that 70% of adults report being online (Fox, 2006), more than one-half of U.S. adults (113 million people) have used the Internet to seek out health information. Of cancer patient populations from academic or community medical centers, 31% to 60% of patients or caregivers report having used the Internet to search for information about cancer (Basch, Thaler, Shi, Yakren, & Schrag, 2004; Metz et al., 2003; Monnier, Laken, & Carter, 2002; Ranson et al., 2003). Clearly, the Internet has significantly transformed the way in which patients and caregivers meet their health-related information needs.
Patients who search for cancer-related Internet information differ considerably from those who do not. Cancer patients who search for information about cancer on the Internet are more likely to be younger, own a computer, have Internet access at home, and have a higher education level than cancer patients who do not search for cancer-related Internet information (Basch, et al., 2004; Helft, Eckles, Johnson-Calley, & Daugherty, 2005; Monnier et al., 2002). In studies of lung and breast cancer patients, education and income were positively correlated with the likelihood of using the Internet to search for cancer-related information (Fogel, Albert, Schnabel, Ditkoff, & Neugut, 2002; Peterson & Fretz, 2003). Although younger age has been correlated with Internet use in prostate cancer and melanoma (Sabel et al., 2005; Smith et al., 2003), this is not the case in breast cancer (Fogel et al., 2002).
Health-related Internet searches can empower patients seeking to become active participants in their care (Fleisher, Bass, Ruzek, & McKeown-Conn, 2002); however, Internet information about treatment options is only empowering to the extent that the oncologist is receptive to the patient being involved in the decision-making process (Broom, 2005b). The Internet provides patients and caregivers access to information that was previously either unavailable or difficult to access, offering a leveling effect on the power imbalance in the doctor–patient relationship, specifically in terms of expert power (Bylund, Sabee, Imes, & Sanford, 2007; Makoul, 1998). It is neither surprising that such a shift affects doctor–patient communication and relationships, nor that 44% of oncologists report having difficulty with discussions about Internet information (Helft, Hlubocky, & Daugherty, 2003).
The percentage of cancer patients who have discussed cancer-related Internet information with their doctors is quite similar to a more general population of patients discussing Internet information (Diaz et al., 2002; Diaz, Sciamanna, Evangelou, Stamp, & Ferguson, 2005; Fox & Rainie, 2002; Murray et al., 2003), with 36% of melanoma patients stating that they intended to discuss Internet information with their physician (Sabel et al., 2005), and 40% of mixed cancer type patients reporting that they had already discussed Internet information with their physician (Helft et al., 2005).
Studies have shown several reasons why discussions about Internet information can be challenging for doctors and patients. Some oncologists perceive Internet-informed patients have a harmful effect on the doctor–patient relationship, particularly when patients directly challenge them or when patients have too much information (Broom, 2005a). After reading cancer-related Internet information, some patients become overwhelmed, aware of conflicting medical information, and more nervous, anxious, and confused (Fleisher et al., 2002; Helft et al., 2005; Sabel et al., 2005). Oncologists also report that Internet information can lead to unnecessary discussion during a consultation that may consequently result in conflict (Broom, 2005a). Internet discussions may result in longer consultations (Helft et al., 2003), which can be frustrating for doctors who have limited time with each patient.
Further, doctors and patients do not always agree about the characteristics of Internet information, which may lead to challenging doctor–patient consultations. For example, cancer patients often make judgments about the accuracy of information that they find on the Internet, and many perceive it to be quite accurate (Helft et al., 2005). However, oncologists report concern over the accuracy of cancer-related Internet information, with 91% of oncologists in one study reporting that the Internet had the potential to cause harm to patients (Broom, 2005a; Newnham et al., 2005). Although 62% of cancer patients in one study reported that cancer-related Internet information made them feel more hopeful (Helft et al., 2005), 38% of oncologists who reported thinking that Internet information makes patients more hopeful viewed this negatively (Helft et al., 2003). These differing perceptions between doctors and patients about cancer-related Internet information may contribute to challenging discussions and may reduce the likelihood of patient-centered communication.
In our previous work, we conceptualized the patient's act of introducing Internet information into a doctor–patient consultation as a face-threatening act (Bylund et al., 2007). How a doctor responds to this information, and the course of the discussion that follows, can prove to either support the patient's face (by validating the patient's efforts or taking the information seriously) or further threaten the patient's face (by warning the patient about the dangers of the Internet or disagreeing with the information without any validation of efforts). Supporting, rather than threatening, the patient's face helps to support two of the key functions of the patient-centered communication model presented in the recent National Cancer Institute monograph: fostering healing relationships and exchanging information (Epstein & Street, 2007). We expect that face-saving or supporting comments and discussion from doctors will lead to better patient outcomes.
In an earlier study about patient communication with health care providers about Internet information in a non-disease-specific patient population, we found that provider disagreement with Internet information resulted in lower patient satisfaction (Bylund, Gueguen, et al., 2007). We also explored which additional strategies might moderate disagreement and found that when a provider disagreed with the Internet information, also taking the information seriously improved patient satisfaction with the interaction.
STUDY AIMS
Researchers have yet to establish effective doctor and patient communication strategies to facilitate these potentially challenging discussions. The purpose of the current study was to examine recalled conversations about Internet information between cancer patients (or caregivers) and their doctors. We focused the current study around three primary research questions.
First, what strategies do patients and caregivers use to introduce Internet information during a consultation with their doctor?
Second, what strategies do doctors use to respond to the Internet information and is the choice of strategy affected by the patient or caregiver's introduction?
Third, are the response strategies linked to outcomes of the conversation?
We hypothesized that in cases where the doctor disagreed with the patient's Internet information, taking the information seriously would improve patient outcomes.
METHOD
Participants
Two hundred and sixty-seven cancer patients and caregivers responded to a questionnaire about their experiences with Internet information about their cancer. Of the 267 survey participants, 238 (89%) reported having searched the Internet for information since the diagnosis of cancer. Of those, 38% reported always talking about the Internet information when they thought it was important; 59% said that they sometimes discussed the information when they thought it was important; and 3% reported never discussing the information, even when they thought it was important. As the current study was focused on participants’ experiences discussing Internet information with their doctors, the sample of interest is the 230 participants who sometimes or always talk about Internet information with their doctors.
Participants were recruited from three online communities sponsored by the National Comprehensive Cancer Network (NCCN), a not-for-profit alliance of 21 leading cancer centers. Participants were recruited via commercially available panels and invitations on NCCN Centers’ Websites. The purpose of hosting these online communities was to field surveys and discussion topics about how patients and caregivers make treatment-related decisions. In total, 900 patients and caregivers participated in these online communities. As a part of their community membership, all participants were asked to respond to a variety of discussion threads and surveys. One of these options was a survey about patients’ and caregivers’ experiences finding information on the Internet about cancer and the discussions they had about this information with their doctors. Although they were recruited from different online communities, each participant received the same survey. The study was reviewed and deemed exempt by our Institutional Review Board.
Table 1 displays the patient demographics from each of the three online communities. The first community of participants included a mix of patients (59%) and primary caregivers of patients (41%) with a broad range of cancer types, all newly diagnosed. The second community focused on late-stage cancer patients and those caring for people with advanced-stage cancer. In this community, 10% of participants were caregivers. The third and final community focused on four types of cancer (breast, prostate, colon/rectal, and lung) and included only patients who had an early-stage diagnosis. Caregivers reported the patient's demographics, not their own. Overall, patients tended to be female (67%–76%) and White (87%–97%). Efforts to recruit more diverse participants were not successful. Caregivers and patients were unrelated in all samples (i.e., both parties of a caregiver–patient dyad were not involved in the study).
TABLE 1.
Demographics
| Group 1 | Group 2 | Group 3 | |
|---|---|---|---|
| Gender | |||
| Male | 26 (32.5) | 18 (23.7) | 19 (25.7) |
| Female | 54 (67.5) | 58 (76.3) | 55 (74.3) |
| Ethnicity | |||
| White/Caucasian | 70 (87.5) | 74 (97.4) | 65 (87.8) |
| Other | 10 (12.5) | 2 (2.6) | 9 (12.2) |
| Age (Group 1) | |||
| 25–44 | 17 (21.2) | a | a |
| 45–64 | 54 (67.5) | a | a |
| ≥65 | 9 (11.3) | a | a |
| Age (Group 2,3) | |||
| 25–39 | a | 5 (6.6) | 17 (23) |
| 40–49 | a | 25 (32.9) | 25 (33.8) |
| 50–64 | a | 38 (50) | 28 (37.8) |
| ≥ 65 | a | 8 (10.5) | 4 (5.4) |
| Education Level | |||
| Some High School | 6 (7.5) | 0 (0) | 0 (0) |
| High School Graduate | 20 (25) | 12 (15.8) | 5 (6.8) |
| Some College | 22 (27.5) | 22 (28.9) | 9 (12.2) |
| College Graduate | 17 (21.2) | 19 (25) | 31 (41.9) |
| Post-Graduate Education | 15 (18.8) | 23 (30.3) | 20 (27) |
| Some Technical or Vocational | 0 (0) | 0 (0) | 9 (12.1) |
| Marital Status | |||
| Married | 54 (67.5) | 55 (72.4) | a |
| Divorced | 9 (11.25) | 9 (11.8) | a |
| Widowed | 9 (11.25) | 4 (5.3) | a |
| Single/Never Married | 4 (5) | 3 (3.9) | a |
| Living with Significant Other | 4 (5) | 5 (6.6) | a |
| Income Level | |||
| < $50,000 | 36 (45) | 18 (23.7) | 15 (20.3) |
| $50,000–$75,000 | 19 (23.75) | 24 (31.6) | 12 (16.2) |
| $75,000–$100,000 | 8 (10) | 14 (18.4) | 22 (29.7) |
| >$100,000 | 14 (17.5) | 11 (14.5) | 19 (25.7) |
| Don't know/Decline to answer /Missing | 3 (3.75) | 9 (11.8) | 6 (8.1) |
| Region | |||
| Midwest | 17 (21.25) | 23 (30.3) | 8 (10.8) |
| Northeast | 22 (27.5) | 13 (17.1) | 24 (32.4) |
| Southern | 23 (28.75) | 28 (36.8) | 26 (35.1) |
| West | 18 (22.5) | 12 (15.8) | 16 (21.6) |
| Time Since Diagnosis | |||
| < 6 mos | 38 (47.5) | 10 (13.2) | 10 (13.5) |
| 6 mos–1 yr. | 23 (28.75) | 11 (14.5) | 23 (31.1) |
| 1 yr.–3 yrs. | 19 (23.75) | 21 (27.6) | 34 (45.9) |
| 3 yrs.–5 yrs. | 0 (0) | 19 (25) | 4 (5.4) |
| >5 yrs. | 0 (0) | 15 (19.7) | 3 (4.1) |
| Cancer Type | |||
| Breast | 26 (32.5) | 28 (36.8) | 43 (58.1) |
| Lung | 9 (11.25) | 10 (13.2) | 9 (12.2) |
| Gynecological | 5 (6.25) | 1 (1.3) | 0 (0) |
| Gastrointestinal | 8 (10) | 15 (19.7) | 16 (21.6) |
| Leukemia/Lymphoma | 7 (8.75) | 1 (1.3) | 0 (0) |
| Genitourinary | 8 (10) | 9 (11.8) | 6 (8.1) |
| Multiple | 0 (0) | 10 (13.2) | 0 (0) |
| Other | 17 (21.25) | 2 (2.6) | 0 (0) |
| Type of Facilityb | |||
| Specialty Oncology Physician | 42 (52.5) | 50 (65.8) | 36 (48.6) |
| Academic/Teaching Hospital | 17 (21.25) | 14 (18.4) | 12 (16.2) |
| Community Hospital | 21 (26.25) | 15 (19.7) | 22 (29.7) |
| NCI Designated Center | 10 (12.5) | 10 (13.2) | 14 (18.9) |
Data not collected.
Multiple responses allowed.
NCI = National Cancer Institute.
Questionnaire
Participants completed an online questionnaire regarding their experiences with searching for cancer information on the Internet and discussing it with their doctors. They were instructed: “Think about the most recent instance since your diagnosis in which you talked with your doctor about cancer information that you found on the Internet. Please keep this situation in mind while answering the following questions.” Using a conversation recall technique (Bylund, Gueguen, et al., 2007), participants were first asked to write down the conversation that they had with the doctor in as much detail as possible. The advantage of the conversation recall technique is that it focuses on participants’ perceptions of the conversation. Although it may not be completely accurate, it does represent what the participant believes she or he said and heard (Lazarus, 1991; Weiner, 1986). Participants were asked several questions about the Internet information that they discussed, including on which Websites the Internet information was found, what types of topics were discussed, if they brought a printed copy of the information with them to the consultation, and if they told the doctor the information was from the Internet.
Patient outcomes
Study participants were also asked to respond to items about the outcomes of the conversation, including (1) a Likert-type, 5-point item measuring satisfaction with the conversation; (2) what impact, if any, the conversation had on choice of doctor or where to seek treatment; and (3) how they would have changed the conversation with the doctor if they could change it.
Coding
The Internet Health Discussion Analysis System (IHDAS) (Bylund, Gueguen, et al., 2007) was used to code recalled conversation data for the strategies used by participants to introduce the information and the providers’ responses. All recalled conversations were coded for each set of codes by the third author. The second author also coded a randomly selected 20% of recalled conversations. Good reliability was established across both sets of codes (Cohen's kappa = .82 for form of introduction and .81 to 1.0 for doctor response codes).
Introduction of Internet information
Consistent with the IHDAS, each recalled conversation was first coded for how the Internet information was introduced by the participant. A code of Facts was given if a statement of fact was made to the doctor without indication that the participant was “owning” the information or actively processing it (e.g., drawing a conclusion, making a decision). The Indirect/Embedded Question code was given in cases where a participant elicited a response from the doctor without using a traditional question form. The Active/Assertive code was given when the information was introduced in such a way that indicated the participant had come to a conclusion or had made a reasoned judgment about diagnosis, treatment, or the next steps that should occur. A code of Direct Question was given when the participant took a more “traditional” patient role by asking a question about the doctor's opinion or knowledge about something.
Doctor responses
Each recalled conversation was also coded for the doctor's response(s) to the Internet health information presented by the participant throughout the entirety of the recalled conversation. There were seven possible codes, and multiple codes could be applied to characterize the responses given in each conversation. Responses included the doctor (1) being Interested/Involved in the Internet information; (2) Actively Agreeing with the participant and/or Internet information; (3) being Dismissive/Argumentative; (4) giving the participant a Warning about Internet information; (5) Taking the Information Seriously; (6) Validating the Participant's Efforts; and/or (7) Disagreeing with the participant and/or Internet information.
Analytic strategies and statistical methods
Descriptive statistics were used to report percentages and averages of the patient and doctor outcomes. Generally, contingency tables analyses using the likelihood-ratio statistics were used to test the association between outcome variables (Wickens, 1989). A special focus was on testing the association between doctors’ responses in the conversation and participants’ recalled outcomes of the conversation. Fisher's exact test was deemed appropriate for testing these associations because of sparse cells in some of the contingency tables. Doctors’ responses were dichotomized (e.g., coded “1” if participant was “taken seriously” by the doctor and “0” otherwise). Participants’ recalled outcomes of the conversation were also dichotomized. For example, satisfaction with the conversation was dichotomized as “very satisfied” versus “less than very satisfied” by the Likert-type ratings. We report odds ratios of 2 by 2 contingency tables, their statistical significance, as well as the corresponding 95% confidence intervals by Fisher's test. Fisher's tests were carried out using the statistical language R (R Development Core Team, 2004).
RESULTS
Sources and Topics of Internet Information
Participants reported that the Internet information they discussed with their doctors came from a wide variety of Internet sites. As reported in Table 2, 54.3% of participants found information on a cancer organization Website (e.g., American Cancer Society) and 31.7% reported using WebMD as their source. As shown in Table 3, the most frequently reported topic was prognosis, accounting for 43.9% of the topics given. The second and third most frequently reported topics were proven or traditional treatments (29.6%) and clinical trials (22.6%). Less than one fourth of participants (23%) reported that they brought a copy of the Internet information to show to their doctor, although 41.7% reported that they specifically mentioned to the doctor that the information was from the Internet. These findings did not differ based on participant community or caregiver/patient status.
TABLE 2.
Sources of Online Information that Participants Discussed Most Recently with Doctors. N = How Many Provided This Informationa
| Online source | n (%) |
|---|---|
| Cancer organization website (e.g., American Cancer Society) | 125 (54.3) |
| WebMD | 73 (31.7) |
| Hospital/Cancer Center website | 56 (24.3) |
| Foundation website (e.g., Susan G. Komen Breast Cancer Foundation) | 54 (23.5) |
| Government website (e.g., NIH, NCI, CDC) | 53 (23) |
| Pharmaceutical company or drug website | 40 (l7.4) |
| Online scientific journal article | 35 (15.2) |
| Health chat room or message board | 28 (12.2) |
| PubMed lit search | 25 (10.9) |
| Other | 23 (10) |
| Comprehensive internet service provider (e.g., AOL, Yahoo, MSN) | 20 (8.7) |
| Media website (e.g., CNN.com) | 16 (7) |
| Individuals' home pages or blog | 11 (4. 8) |
NIH = National Institutes of Health; NCI = National Cancer Institute; CDC = Centers for Disease Control.
Multiple responses allowed.
TABLE 3.
Topics of Online Information that Participants Discussed Most Recently with Doctors. N = How Many People Provided This Informationa
| Topic of Discussion | n (%) |
|---|---|
| Prognosis | 101 (43.9) |
| Proven or Traditional Treatments | 68 (29.6) |
| Clinical Trials | 52 (22.6) |
| Alternative Therapies | 48 (20.9) |
| Diagnosis | 44 (19.1) |
| Experiences of other patients | 41 (17.8) |
| Treatments offered by other MD's or Centers | 40 (17.4) |
| Other | 23 (10) |
Multiple responses allowed.
Introducing the Internet Information
One-hundred and ninety-two participants provided some form of recalled conversation, though only 123 of these recalled conversations were complete enough to be coded using the IHDAS. As shown in Table 4, participants most frequently used a direct question (46.7%), thus putting themselves in a more traditional patient role when introducing the information. The second most frequent form of introduction, accounting for 26.2% of total introductions coded, was simply stating the facts or content of the information without asking a question, followed by active/assertive introductions (22.1%) and indirect/embedded questions (4.9%). Patients (26.3%) were more likely than caregivers (4.2%) to use an assertive strategy to introduce the information (χ 2 = 8.47, p < .05). Self-reported comfort level and frequency of discussing Internet information with the doctor was not associated with the form of introduction.
TABLE 4.
Participant Strategies for Introducing Internet Information in Their Most Recent Discussion with Doctors
| Code Name | Example | n (%) |
|---|---|---|
| Direct question | “What are the chances the cancer will return?” | 57 (46.7) |
| Facts | “According to what I have found on the Internet, the prognosis is up to five years.” | 32 (26.2) |
| Active/ Assertive | “I'm interested in having a new test done to see if I'm metabolizing Tamoxifen.” | 27 (22.1) |
| Indirect/Embedded question | “I am concerned that the cancer might come back.” | 6 (4.9) |
Doctors’ Responses to Internet Information
Doctors were most likely to use the strategy of take the information seriously (82.1%), followed by disagree (20.3%), interested/involved (17.1%) and actively agree (8.9%) (see Table 5). Community number and patient/caregiver status were not associated with doctor response.
TABLE 5.
Doctors' Responses to Patient-Introduced Internet Information
| Doctor Response Codes | Example | n (%) |
|---|---|---|
| Take the information seriously | “Let's wait until a month or so has passed and do a few more tests to determine if hormone treatment is necessary.” | 101 (82.1) |
| Disagree | “Absolutely not.” | 25 (20.3) |
| Interested/Involved | “I will research it and see if it is something that you would be able to use.” | 21 (17.1) |
| Actively agree | “Yes, that would be a good idea. You can talk to him or we can refer you to a surgeon.” | 11 (8.9) |
| Warning | “Don't believe everything you read on the Internet.” | 5 (4.1) |
| Dismissive / Argumentative | “That's not really going to do anything. We need to stay focused on the treatment you are already on.” | 4 (3.3) |
| Validating patient efforts | “You're a lot more up to date than most of my patients.” | 3 (2.4) |
To address the second research question, we examined the association between the manner in which the participant introduced the information and the doctor's response. Doctors’ responses were more often coded as interested/involved when the participant told the doctor that the information was from the Internet (χ 2 = 6.17, p < .05). Second, and in line with our previous work (Bylund, Gueguen, et al., 2007), we created a dichotomous variable of participant introduction forms (as coded by the IHDAS) into less face-threatening (facts, indirect and direct questions) and more face-threatening (assertive). Doctors were more likely to use the strategy of actively agree when participants introduced the information in an assertive manner than a nonassertive manner (χ 2 = 12.25, p < .001).
Effect of Doctor's Response on Participant-Reported Outcomes
We explored if certain physician response strategies were associated with participant-reported outcomes in answer to the third research question. Due to the low frequency of three of the strategies, we used the following four strategies as independent variables: interested/involved, actively agree, take the information seriously, and disagree. For these analyses, we focused on the 123 participants who recalled conversations that we were able to code.
We used three dichotomous participant outcome variables: very satisfied/less than very satisfied, positive impact/not a positive impact, and desire to change the doctor's response/no desire to change based on the measures described above. More than one half (61.2%) of participants reported being “very satisfied” with the conversation they had with the doctor about Internet information. Regarding impact, 22.7% of participants reported the conversation had a positive impact. Finally, 70.7% of responses to the question of if the participant wished he or she could change how the doctor responded were coded as desire to change.
Figure 1 displays the odds ratio of each of the three outcomes when each strategy was present. When the doctor's response was coded as being interested/involved, participants were significantly less likely to report a desire to change the doctor's response (95% Confidence Interval [CI]: .00–.65, p < .01), but there was no significant findings for the outcomes of very satisfied or positive impact. There were also no significant associations with any of the three outcomes when the doctor actively agreed with the patient or caregiver. When the doctor took the information seriously, participants were significantly less likely to report a desire to change the doctor's response (CI: .09–.75, p < .01), and they were also more likely to report high satisfaction (CI: 2.96–43.85, p < .001), though there was no significant finding for positive impact. Finally, when the doctor disagreed with the participant, the participant was more likely to report wanting to change the doctor's response (CI: 1.28–9.71, p < .05) and less likely to report being very satisfied (CI: .04–.42, p < .001).
FIGURE 1.
Odds ratios of response strategy on patient outcome.
Further, we were interested in testing the hypothesis that when a doctor disagreed with the information that the participant brought in, the strategy of take it seriously would improve reported outcomes. In cases where a doctor disagreed but took the participant's information seriously, Table 6 shows that 2 of 8 participants (25%) were “very satisfied” with the consultation. When the doctor did not take it seriously, 4 of 17 participants (24%) remained “very satisfied” with the consultation. By contrast, the strategy of take it seriously made a substantial difference when a doctor did not disagree; 68 of 93 participants (73%) were “very satisfied” when the doctor took it seriously, whereas none of the five participants were “very satisfied” when the doctor did not take it seriously. To better explicate the potential moderating effect of taking it seriously on disagreement, we used the bias-corrected log odds ratio as a measure of association between the “very satisfied” outcome and doctors’ taking it seriously (Wickens, 1989). An odds ratio of 1.0 (equivalently, a log odds ratio of 0.0) represents a null association. In cases where the doctor disagreed with the information, the log odds ratio was close to null, at 0.14 . However, in cases where a doctor did not disagree, the log odds ratio was substantially higher, at 3.39 . The potential moderating effect of take it seriously on the association between disagreement and a “very satisfied” outcome can be tested by fitting a conditional independence log-linear model (Wickens, 1989). The model tests the null hypothesis that the outcome is associated with disagreement across all levels of take it seriously, which was rejected by the likelihood-ratio test statistic (G2 = 12.47, df = 2, p < 0.002). The association between the outcome and disagreement depends on the strategy of take it seriously, although the association was likely attributable mostly to cases where a doctor did not disagree. We were unable to test the interaction between interested/involved and disagree as there were no conversations coded with both of these strategies present.
TABLE 6.
Participant Outcome by Doctor Response
| Disagree = No |
Disagree = Yes |
||||
|---|---|---|---|---|---|
| Serious |
Serious |
||||
| No | Yes | No | Yes | ||
| Patient outcome | not “very satisfied” | 5 | 25 | 13 | 6 |
| “very satisfied” | 0 | 68 | 4 | 2 | |
DISCUSSION
The results of the current study support the importance of patient-centered communication in discussions about Internet information. Two patient-centered communication strategies had significant effects on participant-reported outcomes. When doctors communicated interest and involvement about the Internet information or showed that they took the information seriously, participants were less likely to report a desire to change the doctor's response. Further, when the doctor took the information seriously, participants reported high satisfaction and were less likely to report a desire to change the doctor's response. Showing interest and involvement and taking information seriously in this context can be important to fulfilling at least two key components of patient-centered communication as presented in the NCI monograph (Epstein & Street, 2007). Doctors who use these strategies improve the function of exchanging information as they are able to understand more about the information the participant has. As pointed out in Epstein and Street (2007), “The wide availability of health information on the Internet can cause patients to become frustrated by information ‘overload’ and their inability to clearly understand their health status and appropriate course of treatment. Clinicians must help patients and caregivers filter and understand clinical information” (p. 21). Being interested/involved and taking information seriously is key to achieving this. Further, by using these strategies, doctors are giving a message to the patient or caregiver that their role as an active participant in clinical encounters is important, helping to foster healing relationships.
Results of the current study demonstrated that when the doctor disagreed with the information that the participant raised, participant-reported outcomes were significantly worse than if the doctor did not disagree. Understandably, at times doctors may disagree with information patients or caregivers have found on the Internet—it may not be relevant to the patient's situation or it may not be evidence based. We expected that taking the information seriously or showing interest and involvement would help to buffer disagreement by improving patient outcomes; however, this was not supported. This may have been due to the small sample size as there were just eight cases of doctors disagreeing and showing interest/involvement.
The way that the patient or caregiver introduces the Internet information was shown here to affect the doctor's response, demonstrating the mutual effect of communication (Cegala, Street, & Clinch, 2007). When patients or caregivers explicitly stated that the information was from the Internet, doctors were more likely to demonstrate interest and involvement. When patients introduced information in an active and assertive way, they were more likely to garner the response of actively agree from the doctor. In these cases, doctors may have felt threatened and a need to appease the patient (Bylund, Gueguen, et al., 2007). This latter relationship was one we also found in our earlier work with a non-cancer-specific population (Bylund, Gueguen, et al., 2007). In that study, as in this one, we also found disagreement to be related to lower patient satisfaction and taking information seriously to be associated with higher patient satisfaction. The ability to replicate findings from a different sample focused specifically on cancer patients indicates externally valid findings.
Limitations of the current study include the relative homogeneity of participants, its reliance on self-report data, and that we only were able to get the patient's perspective of the doctor–patient encounter. We are currently conducting an observational study wherein we are obtaining audio-recorded conversations between patients and doctors about Internet information and capturing doctor and patient self-report data. Other future work should further explore the way in which these conversations can be improved, particularly in cases where the information the patient has found is incorrect and misunderstandings need to be corrected. Ultimately, interventions to improve patients’ Internet searching and conversations should be examined for their effects on patient and doctor outcomes.
CONCLUSION
The Internet has changed the way that patients and their caregivers gain information about their cancer and has added a level of complexity to doctor–patient interactions. Although many patients use the Internet to search for cancer-related information, they are not always forthcoming with their doctors about the source of the information they have found, partially as a result of concern over their doctors’ possible response to it. Furthermore, patients use a variety of strategies to introduce the information to their doctors which in turn affects their doctor's response. This preliminary evidence that the doctor's response is associated with patient outcomes indicates the potential for improving patient-centered communication.
CLINICAL IMPLICATIONS.
The way doctors respond to their patients’ presentation of Internet information affects patient satisfaction and whether patients want to change elements of their encounters and their health care decisions. In an effort to maximize patient-centered communication, doctors should be encouraged to take their patients and the information they present seriously, as well as show their patients that they are interested and involved. There is potential for communication skills interventions to have an integral role for doctors to focus on and practice strategies related to this topic. Because patients who present information in an active/assertive way tend to elicit agreeable responses from their doctors, there may also be implications for communication skills interventions for patients focusing on how best to engage doctors in conversation about Internet information they have found that they wish to discuss.
Acknowledgments
This research was supported by Memorial Sloan-Kettering Cancer Center and the National Comprehensive Cancer Network. All data analysis and decisions about publication were made by the authors. This publication acknowledges Grant Number NCI P30 CA08748, which provides partial support for Behavioral Research Methods Core used in conducting this investigation.
Footnotes
The authors have no conflicts of interest to report.
REFERENCES
- Basch E, Thaler H, Shi W, Yakren S, Schrag D. Use of information resources by patients with cancer and their companions. Cancer. 2004;100:2476–2483. doi: 10.1002/cncr.20261. [DOI] [PubMed] [Google Scholar]
- Broom A. Medical specialists’ accounts of the impact of the Internet on the doctor/patient relationship. Health. 2005a;9(3):319–338. doi: 10.1177/1363459305052903. [DOI] [PubMed] [Google Scholar]
- Broom A. Virtually healthy: The impact of Internet use on disease experience and the doctor-patient relationship. Qualitative Health Research. 2005b;15(3):325–345. doi: 10.1177/1049732304272916. [DOI] [PubMed] [Google Scholar]
- Bylund CL, Gueguen JA, Sabee CM, Imes RS, Li Y, Sanford AA. Provider-patient dialogue about Internet health information: An exploration of strategies to improve the provider-patient relationship. Patient Education and Counseling. 2007;66:346–352. doi: 10.1016/j.pec.2007.01.009. [DOI] [PubMed] [Google Scholar]
- Bylund CL, Sabee CM, Imes RS, Sanford AA. Exploration of the construct of reliance among patients who talk with their providers about Internet information. Journal of Health Communication. 2007;12(1):17–28. doi: 10.1080/10810730601091318. [DOI] [PubMed] [Google Scholar]
- Cegala DJ, Street RL, Clinch CR. The impact of patient participation on physicians’ information provision during a primary care medical interview. Health Communication. 2007;21(2):177–185. doi: 10.1080/10410230701307824. [DOI] [PubMed] [Google Scholar]
- Diaz JA, Griffith RA, Ng JJ, Reinert SE, Friedmann PD, Moulton AW. Patients’ use of the Internet for medical information. Journal of General Internal Medicine. 2002;17:180–185. doi: 10.1046/j.1525-1497.2002.10603.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Diaz JA, Sciamanna CN, Evangelou E, Stamp MJ, Ferguson T. Brief report: What types of Internet guidance do patients want from their physicians? Journal of General Internal Medicine. 2005;20(8):683–685. doi: 10.1111/j.1525-1497.2005.0115.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein R, Street RL., Jr. Patient-centered communication in cancer care: Promoting healthing and reducing suffering. National Cancer Institute; Bethesda, MD: 2007. NIH Publication No. 07-6225. [Google Scholar]
- Fleisher L, Bass S, Ruzek S, McKeown-Conn N. Relationships among Internet health information use, patient behavior and self-efficacy in newly diagnosed cancer patients who contact the National Cancer Institute's (NCI) Atlantic Region Cancer Information Service (CIS). Paper presented at the American Medical Informatics Association (AMIA) Annual Fall Symposium; San Antonio, TX.. Nov, 2002. [PMC free article] [PubMed] [Google Scholar]
- Fogel J, Albert SM, Schnabel F, Ditkoff BA, Neugut AI. Use of the Internet by women with breast cancer. Journal of Medical Internet Research. 2002;4(2):e9. doi: 10.2196/jmir.4.2.e9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fox S. Online health search 2006. Pew Internet & American Life Project; Washington, DC: 2006. [Google Scholar]
- Fox S, Rainie L. Vital decisions: How Internet users decide what information to trust when they or their loved ones are sick. Pew Internet & American Life Project; Washington, DC: 2002. [Google Scholar]
- Helft PR, Eckles RE, Johnson-Calley CS, Daugherty CK. Use of the Internet to obtain cancer information among cancer patients at an urban county hospital. Journal of Clinical Oncology. 2005;23:4954–4962. doi: 10.1200/JCO.2005.09.621. [DOI] [PubMed] [Google Scholar]
- Helft PR, Hlubocky F, Daugherty CK. American oncologists' views of Internet use by cancer patients: A mail survey of American Society of Clinical Oncology members. Journal of Clinical Oncology. 2003;21(5):942–947. doi: 10.1200/JCO.2003.08.007. [DOI] [PubMed] [Google Scholar]
- Lazarus RS. Emotion and adaptation. Oxford University Press; New York: 1991. [Google Scholar]
- Makoul G. Perpetuating passivity: Reliance and reciprocal determinism in physician-patient interaction. Journal of Health Communication. 1998;3:233–259. doi: 10.1080/108107398127355. [DOI] [PubMed] [Google Scholar]
- Metz J, Devine P, DeNittis A, Jones H, Hampshire M, Goldwein J, et al. A multi-institutional study of Internet utilization by radiation oncology patients. International Journal of Radiation Oncology—Biology—Physics. 2003;56(4):1201–1205. doi: 10.1016/s0360-3016(03)00407-3. [DOI] [PubMed] [Google Scholar]
- Monnier J, Laken M, Carter C. Patient and caregiver interest in Internet-based cancer services. Cancer Practice. 2002;10(6):305–310. doi: 10.1046/j.1523-5394.2002.106005.x. [DOI] [PubMed] [Google Scholar]
- Murray E, Lo B, Pollack L, Donelan K, Catania J, White M, et al. The impact of health information on the Internet on the physician-patient relation-ship. Archives of Internal Medicine. 2003;163:1727–1734. doi: 10.1001/archinte.163.14.1727. [DOI] [PubMed] [Google Scholar]
- Newnham G, Burns W, Snyder R, Dowling A, Ranieri N, Gray E, et al. Attitudes of oncology health professionals to information from the Internet and other media. Medical Journal of Australia. 2005;183(4):197–200. doi: 10.5694/j.1326-5377.2005.tb06996.x. [DOI] [PubMed] [Google Scholar]
- Peterson M, Fretz P. Patient use of the Internet for information in a lung cancer clinic. CHEST. 2003;123:452–457. doi: 10.1378/chest.123.2.452. [DOI] [PubMed] [Google Scholar]
- Ranson S, Morrow G, Dakhil S, Wade J, Fitch T, Roscoe J, et al. Internet use among 1020 cancer patients assessed in community practices: A URCC CCOP study. Proceedings/Annual Meeting of the American Society of Clinical Oncology. 2003;22:534. Abstract 2149. [Google Scholar]
- R Development Core Team . R: A language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria: 2004. [Google Scholar]
- Sabel MS, Strecher VJ, Schwartz JL, Wang TS, Karimipour DJ, Orringer JS, et al. Patterns of Internet use and impact on patients with melanoma. Journal of the American Academy of Dermatology. 2005;52:779–785. doi: 10.1016/j.jaad.2004.10.874. [DOI] [PubMed] [Google Scholar]
- Smith R, Devine P, Jones H, DeNittis A, Whittington R, Metz J. Internet use by patients with prostate cancer undergoing radiotherapy. Urology. 2003;62:273–277. doi: 10.1016/s0090-4295(03)00251-6. [DOI] [PubMed] [Google Scholar]
- Weiner B. An attributional theory of motivation and emotion. Springer-Verlag; New York: 1986. [Google Scholar]
- Wickens TD. Multiway contingency tables analysis for the social sciences. Erlbaum Associates; Hillsdale, NJ: 1989. [Google Scholar]