Patients' Preferences for Biopsy Result Notification in an Era of Electronic Messaging Methods

Importance  Effective patient-physician communication is essential for optimal health care. Recent introduction of online patient portals to access test results are changing the communication landscape, but regulatory guidelines for the online release of biopsy results vary from state to state.

Objectives  To assess patient preferences for receiving skin biopsy results to rule out melanoma and to compare those preferences to current physicians’ practices for notification.

Design, Setting, and Participants  English-speaking individuals 18 years or older were recruited consecutively from melanoma clinics at 3 academic tertiary referral medical centers: University of California, San Francisco, University of Pennsylvania, and Duke University. Patients were surveyed from July 1, 2012, through July 31, 2013. A second survey of physicians at these institutions was conducted to assess physician notification practices.

Results  A total of 301 of 305 patients agreed to participate (98.7 response rate). Most of the patients (67.1%) preferred to speak directly with their physician by telephone to receive their skin biopsy results, followed by a distant second choice (19.5%) of being notified in person at a clinic visit. Voice message or online patient portal were each the preferred method of communication for 5.1% of patients. The most important consideration for patients was a communication modality that provided test results in the most rapid manner; 51.7% wanted a method that was rapid, and 7.8% preferred a method that was not only speedy but also allowed them an opportunity to ask questions. A total of 59.5% of the study participants would choose the same communication method regardless of the biopsy results, but 40.5% preferred a different mode of notification if their biopsy results revealed a malignant tumor. Younger and more highly educated patients favored the online portal. Of 84 physicians surveyed, 47 responded (56% response rate). Physicians’ overall preferred method of contacting patients aligned with patient preference for speaking by telephone (56.5%). However, for benign results, 31.2% of physicians chose to speak by telephone, whereas patients preferred voicemail (32.1%). There was physician uncertainty as to guidelines regarding communication of test results.

Conclusions and Relevance  Patient preference has shifted from face-to-face visit to discussion over the telephone because of a desire for rapid notification. Experience with online portal delivery of results favorably inclined patients toward that modality. We recommend that patients be queried regarding their notification preference on the biopsy consent form.

Effective patient-physician communication is integral to providing quality health care and improves patient outcomes.¹ Poor communication of results can lead to patient stress and safety issues.² Up to one-third of physicians may not notify patients of abnormal test results because of reasons such as the benign nature of the results, upcoming scheduled patient clinic visits, and forgetfulness.³ However, the “no news is good news” approach does not meet patient expectations even for normal results.^3-5 Patients have increased anxiety after a skin biopsy, which decreases once they receive the results, regardless of whether the diagnosis is benign or malignant.⁶

The specter of a malignant diagnosis, such as melanoma, can cause patients significant psychological distress.⁷ Previous studies^8-11 of patients with cancer indicate that patients prefer to receive bad news in person, almost never by telephone, and as soon as the diagnosis is confirmed, but a clinic visit can be in conflict with the desire for more rapid communication of the diagnosis. Patients value timeliness of receiving results because delays can lead to increased anxiety.^2,5,6,12-14 Current guidelines on how to break bad news to patients with cancer are surprisingly not based on any systematic studies.^15-18

Test results can be communicated face-to-face, by telephone, or via a secure electronic messaging system, email, or text message; each method has its own advantages and disadvantages.¹⁹ A face-to-face visit is often considered the standard for discussion of information²⁰; however, a telephone call is associated with a high level of patient satisfaction^5,21 and is often preferred for abnormal results.^22,23 However, there is considerable variability in preferences, and some studies^22-24 have found discrepancies between communication methods practiced by physicians and those preferred by patients.

In the past, most patients preferred being told face-to-face about a possible melanoma diagnosis¹²; those informed over the telephone would have preferred to be told in person.²⁵ Newer electronic methods, such as secure online portals, have the potential to change the communication landscape. The use of electronic health records with patient portal functionality has increased rapidly, in part because of federally mandated penalties on physicians who do not demonstrate meaningful use of a qualified electronic health record by 2015.²⁶ Online portals provide information rapidly, without the security concerns that patients have with email,^5,27-30 but the regulations that govern electronic release of reports have no uniformity across states.³¹

Our study assesses patients’ preferences for the method of receiving skin biopsy results to rule out melanoma, collects data on the reasons for their preferences, and examines the degree of concordance between patient preference and physician practice. We identify subgroup preferences that differ from the consensus preferences and conclude with a discussion of the effects of a changing regulatory landscape on physician practice and provide recommendations for a more effective alignment of physician practice with patient preferences.

This study was conducted at the pigmented lesion or melanoma surveillance clinics at 3 academic tertiary referral medical centers: University of California, San Francisco (UCSF), University of Pennsylvania (UPenn), and Duke University (Duke). Institutional review board approval was obtained at each center, and written informed consent was obtained from all patients. Patients were recruited consecutively as they arrived in clinic if they were 18 years or older and English speaking. A previous history of melanoma or nevus was not used as an eligibility criterion. Patients were surveyed during a 12-month period from July 1, 2012, through July 31, 2013. The survey yielded a 98.7% response rate; 301 of 305 patients agreed to participate. Patients completed the questionnaire while waiting to be seen by their physician (ie, before the clinic visit started). At the time of filling out the questionnaire, they were unaware whether a biopsy would be needed at that visit. The CONSORT diagram shows participant flow throughout the study (Figure 1).

Patients were asked to indicate the purpose of the visit, to provide demographic information, and to rank their preferred method of being notified of a biopsy result in order of preference. Patients were asked to self-report race/ethnicity from a classification list provided to assess whether race/ethnicity influenced preferences. Patients were also asked to rank, in order of preference, the reasons for selecting a specific notification modality; if none of the offered choices was sufficient, space was provided where other reasons could be listed. Patients were asked if they preferred a different notification method for biopsy results depending on whether the results were normal vs abnormal. Those who preferred different methods were asked to rank their communication preferences for normal and abnormal results independently. The survey instrument is included in eFigure 1 in the Supplement.

Physicians at the 3 sites were surveyed using the UCSF Research Electronic Data Capture anonymous online questionnaire. The 47 (of 84 total) physicians who completed the survey were asked to provide demographic information and to score various methods of relaying test results to patients when the results were normal vs abnormal. Participants were asked whether they were aware of any departmental or university policy for notifying patients of skin biopsy results.

Demographic characteristics, including age, ethnicity, sex, educational level, and distance from clinic, were compared across the 3 sites included in the study (Table 1). Normally distributed continuous variables were compared using analysis of variance tests; nonparametric Kruskal-Wallis tests were used for those not normally distributed. Categorical variables were compared using χ² tests.

Logistic regression analyses were used to compare a preference for nonelectronic communication (face-to-face visit, telephone, and voice message) compared with electronic communication (email, secure online portal, text message, and private tweet); because a number of patients expressed different preferences depending on whether biopsy results were normal or abnormal, these diagnoses were examined separately. Odds ratios and 95% CIs were calculated for each predictor, which were modeled individually, adjusted for site. For the purposes of the analysis, any question that was not answered correctly according to the instructions provided in the survey was excluded from the statistical analysis, even though the other questions from the same patient’s survey may have been included. As a result, the number of respondents to each item varies slightly; the total number of respondents for whom any item was included in the analysis was 298. The number of patients who ranked a specific communication method as their top choice (scored 1 on a scale of 1 to 7) was summarized for each modality.

The numbers of participants at UCSF, UPenn, and Duke were 116 (38.9%), 108 (36.2%), and 74 (24.8%), respectively. Across our 3 sites, the participants did not differ significantly in terms of their baseline characteristics except in stage of disease and distance from the clinic (Table 1). Mean age of the participants was 54 years; 46.6% were men, 96.3% were white, 88.6% had a previous history of melanoma, and 35.6% had a family history of melanoma. With respect to the highest level of education achieved, participants in our study had the following distribution: 34.2% went to graduate schools, 47.7% went to college, 16.8% completed high school, and 1.3% did not have a high school diploma. Stage I melanoma was the most prevalent (52.0%), followed by stage 0, also known as melanoma in situ (18.1%).

Most of our study cohort (67.1%) preferred to speak directly with their physician by telephone to receive their skin biopsy results, followed by a distant second choice (19.5%) of being notified in person at a clinic visit (Figure 2A). Voice message or online patient portal were each chosen as the preferred method of communication by 5.1% of patients. A total of 2.7% of patients chose email, 0.3% preferred text message, and none selected private tweet. Results for UCSF and UPenn closely mirrored the overall primary analysis and were closely matched, despite statistically significant differences in patient distance from the clinic and tumor staging, whereas the answers from patients at Duke differed in some aspects. A total of 14.1% of the participants at Duke indicated that a secure message via an online patient portal was their preferred method for communication, making it the second ranked item at Duke, after telephone call (Figure 2B), whereas only 2.8% of participants at UPenn and 1.8% participants at UCSF did the same. Of interest, face-to-face communication, which was the second choice at UCSF (26.5%) and UPenn (21.3%), was preferred by only 5.6% at Duke. Duke patients’ enthusiasm for the online patient portal resulted in voice message and online patient portal being tied for third position in the combined analysis.

The most important consideration for patients was a communication modality that provided test results in the most rapid manner; 51.7% wanted a method that was rapid, and 7.8% preferred a method that was not only speedy but also allowed them an opportunity to ask questions. A total of 19.5% of the study cohort thought that meeting with the physician in person was very important, and 8.8% stated that even though they would prefer to meet with the physician, they lived too far away from the clinic to be able to do so. A total of 5.4% preferred an interactive method, 2.0% preferred an electronic method, and 3.4% gave alternative reasons that did not fall under any of the aforementioned categories.

More than half of the study participants (59.5%) would choose the same communication method regardless of whether the biopsy results were normal or abnormal. However, a substantial number of participants (40.5%) stated that they would prefer a different mode of notification if their biopsy results were abnormal. In the case of normal biopsy results, 32.1% of these patients preferred voice message, 25.7% chose email, 23.9% favored speaking by telephone, 14.7% favored the online patient portal, 2.8% preferred text message, 0.9% chose clinic visit, and none chose private tweet (Figure 3). When the results were abnormal, the preferences were similar to those who chose the same communication method for normal or abnormal results.

No significant difference in participant preferences was found based on age, sex, educational level, ethnicity, stage, previous history of malignant melanoma, and traveling distance from clinic (eFigures 2-7 in the Supplement). Speaking by telephone and notification in person at a clinic visit remained the leading preferences in all categories. However, we observed that 2 specific groups (the youngest [18-35 years of age] and college-educated participants) had the most favorable opinion for notification via an online patient portal. Logistic regression analyses were conducted to examine potential predictors of a preference for notification by nonelectronic methods (face-to-face visit, telephone, and voice message) compared with electronic communication methods (email, secure online portal, text message, and private tweet) for normal vs abnormal results independently. Younger and more educated patients favored electronic methods when their biopsy results were normal; however, no difference was found in patient preferences when biopsy results were abnormal (Table 2).

Of the 84 physicians surveyed, 47 responded, for an overall response rate of 56.0%. Mean physician age was 43 years, 59.6% were women, and the mean duration in practice was 11 years. When a patient’s biopsy results were positive for melanoma, physicians preferred telephone call (56.5%) as their top choice, 12.9% would forward a message to an assistant to contact the patient, 9.8% would write a letter, 8.2% would set up an appointment, 5.3% would use a secure online portal, 4.3% would leave a voice message, and 2.4% would send an email. If the results were negative for melanoma, 31.2% would call patients to notify, 23.4% would forward a message to an assistant to contact the patient, 14.4% would leave a voice message, 12.9% would send a letter, 8.9% would use a secure online portal, 6.0% would set up an appointment, and 2.0% would send an email. Most physicians said that they were not aware of any departmental (89.4%) or university policy (97.9%) for biopsy result notification. The few dermatologists who said that they were aware of such a policy had no uniformity as to the method they thought was approved by their department or university.

Our survey revealed that patient preference markedly favored speaking directly with the physician by telephone because of the ability to discuss results with a physician as soon as possible after the results are finalized. This finding is a shift from patients’ previously reported primary preference for a face-to-face clinic visit.^12,25 We found that a face-to-face visit ranked as a distant second choice. The responses from Duke’s participants differed in some significant aspects from the other 2 sites because they ranked secure online portal second, after telephone call; this finding correlates with the Duke patient cohort’s longer experience and familiarity with receiving biopsy results via online portal. The leading reason given for selecting a communication modality was speed of notification. Physician preferences were generally in alignment with patient preference for speaking by telephone for abnormal results, except that for benign results, physicians also chose to speak by telephone, whereas patients preferred voicemail.

Our study population had characteristics that differ from the general population. In our study, women outnumbered men (53.4% vs 46.6%); in contrast, in the United States, more men (58%) are diagnosed as having melanoma than women.³² Women have higher rates of self-detection of moles and melanoma³³ and when evaluated for suspicious lesions are much more anxious than men.⁶ A diagnosis of melanoma can lead to significantly more anxiety and distress among women,^7,34 and women are more likely than men to visit a melanoma clinic at a tertiary care hospital.³⁵ Our study participants also have a higher level of education than the general population. In the United States, 30% have a college degree or higher³⁶ compared with 81.9% in our study, likely because UCSF and Duke are located in metropolitan areas with a higher percentage of college-educated adults than the national metropolitan average³⁷ and because previous work³⁸ has found that highly educated patients have better access to health care and that patients with melanoma at a university-affiliated teaching hospital were generally more well educated and had higher socioeconomic status than patients with melanoma at a nearby public hospital.³⁹

Duke participants indicated that the online portal option was their second most preferred method; this finding correlates with these patients’ longer duration of exposure to and experience with this modality. Duke has been releasing biopsy results via patient online portal since 2009. Both UCSF and UPenn do not release biopsy results online because of state-specific restrictions. California law prohibits electronic sharing of biopsy results if they are positive for a malignant tumor, unless a health care professional has first discussed them with the patient by means of oral communication. Pennsylvania law prohibits direct release of test results to patients, and North Carolina has no laws on the issue.^31,40,41 New federal regulations will supercede all state laws and require clinical laboratories to provide patients with direct access to test reports within 30 days of their request.⁴¹ Because the rapidity of receiving results is paramount to patients in our and other studies,^2,5,6,12-14 and the stated need for interactive discussion of results is secondary, it is possible that notification of biopsy results by online portal will overtake telephone and clinic visit as the method preferred by patients for biopsy notification in the future.

Accessing complex results via online portals may increase patient stress because of misinterpretation in the absence of counseling from a physician. However, data suggest that patients do not experience negative emotional reactions when viewing test results on patient portals without the benefit of physician interpretation.^42,43 For example, a study⁴³ of patients with breast cancer found that having access to imaging and test results on a patient portal resulted in reduced anxiety scores. Another concern is that patients accessing their results online directly may increase telephone calls to physicians and significantly increase physician workload. Studies^28,44 indicate that electronic communication reduces the number of telephone calls to physicians, and over time the amount of time spent answering messages and the volume of messages are reduced. A further concern is that patient queries may be directed to the incorrect physician, for example, to the physician interpreting the result, such as the pathologist, rather than directed to the physician who provided care and ordered the study. For instance, since the online release of radiologic results at UPenn, interpreting radiologists have started receiving telephone calls directly from patients (E.C., unpublished data, June 2014). A solution would be to include information regarding whom to contact regarding the results.

Logistical regression analysis revealed that younger and more educated patients favored electronic methods, such as patient portals and email, when their biopsy results were normal. Even though our study did not find any difference in preferences for electronic vs nonelectronic methods between whites and ethnic minorities, this finding is based on a small number of minority patients (3.7%) who were English speaking. Previous studies^45,46 have found that older patients, those with less education, and ethnic minorities have limited Internet access and lower computer literacy, resulting in infrequent enrollment and use of online patient portals. Studies have suggested that use of aggressive marketing strategies⁴⁵ and targeted dissemination via ethnic media may help offset the digital divide.⁴⁶ Some digital disparities may be mitigated by the ubiquity of mobile telephones in this country. More Americans own mobile telephones than computers (91% vs 76%); more than half of these devices are smartphones that function as handheld computers.⁴⁷ Two-thirds of US cell phone owners use their telephones for web browsing, including 75% of African American and 68% of Hispanic telephone owners.⁴⁸ Online patient portals at all 3 of our participating institutions are accessible on smartphones. However, research has found that limited English proficiency is a significant barrier to accessing health care⁴⁹; efforts could be directed toward increasing portal availability in alternative languages and increasing e–health literacy among disadvantaged groups.

Our study has some limitations, including a small sample size and a focus on academic tertiary referral medical centers. Responses from patients seen at such centers may not be generalizable to those of patients seen in smaller community settings. Ethnic minorities were not well represented in our study, and the patients who were included were English speaking. Responses from this small sample may not be representative of the preferences of non–English-speaking ethnic minorities. Although we found no significant differences between the responses of patients with and without a history of melanoma, we did not query the patients about the number of prior biopsies they had had and the means by which they were notified of the results. Their prior experiences with skin biopsy result notification may have influenced their current response.

The results of our study are likely relevant to biopsies performed by other specialists. Patients are not often asked about their preference for test result notification.⁵ We recommend asking about patients’ notification preference by including this question on the biopsy consent form. We also recommend developing notification guidelines for physicians for both normal and abnormal test results in concordance with state and federal regulations because of physician uncertainty regarding departmental and university policies for biopsy result notification.

Accepted for Publication: December 21, 2014.

Corresponding Author: Maria L. Wei, MD, PhD, Department of Dermatology, University of California, San Francisco, 4150 Clement St, Mailstop 190, San Francisco, CA 94121 (maria.wei@ucsf.edu).

Published Online: April 1, 2015. doi:10.1001/jamadermatol.2014.5634.

Author Contributions: Drs Choudhry and Wei had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Nelson, Wei, Nguyen.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Choudhry, Hong, Wei, Nguyen.

Critical revision of the manuscript for important intellectual content: Choudhry, Chong, Jiang, Hartman, Chu, Nelson, Wei, Nguyen.

Statistical analysis: Choudhry, Wei.

Administrative, technical, or material support: Choudhry, Chong, Nelson, Nguyen.

Study supervision: Nelson, Wei, Nguyen.

Conflict of Interest Disclosures: None reported.

Additional Contributions: Nancy Hills, PhD (UCSF), provided excellent assistance with the statistical analysis. Neil Prose, MD (Duke), provided invaluable discussions and contributed to the conceptualization. Dr Hills was compensated for her time.