Providing Hospital Patients with Access to Their Medical Records

Abstract

Being a hospital patient can be isolating and anxiety-inducing. We conducted two experiments to better understand clinician and patient perceptions about giving patients access to their medical records during hospital encounters. The first experiment, a survey of physicians, nurses, and other care providers (N=53), showed that most respondents were comfortable with the idea of providing patients with their clinical information. Some expressed reservations that patients might misunderstand information and become unnecessarily alarmed or offended. In the second experiment, we provided eight hospital patients with a daily copy of their full medical record—including physician notes and diagnostic test results. From semi-structured interviews with seven of these patients, we found that they perceived the information as highly useful even if they did not fully understand complex medical terms. Our results suggest that increased patient information sharing in the inpatient setting is beneficial and desirable to patients, and generally acceptable to clinicians.

Introduction

Being in the hospital has been called “one of the most dis-empowering situations one can experience in modern society.”¹ Patients often feel isolated, anxious, and that they do not have control over their care. They often do not know what medications they are taking, their treatment plans, or even the names of members of their care teams.^2–4 Therefore, it is important to try to engage hospital patients and facilitate greater participation in their care.

Engaging patients has received significant attention in recent years. In contrast to the traditionally paternalistic doctor-patient relationship, consumers increasingly expect direct and fast access to their health records.⁵ This engagement of patients not only helps the patients feel more involved, but also can lead to improved health outcomes.^6,7 The journal Health Affairs recently referred to patient engagement as equivalent to the next “blockbuster drug”.⁸ The Institute of Medicine (IOM) recommended that individuals receive opportunities to have access to medical information and clinical knowledge which will enable the patients to be the “source of control”.⁹ The U.S. Meaningful Use financial incentive program stipulates requirements for patients having access to clinical summaries, electronic messaging with their providers, patient-specific educational resources and online access to their personal health information (including information like care team members, medication lists and history, laboratory results, and problem lists).¹⁰ Further, the Consumer Assessment of Healthcare Providers and Systems (CAPHPS) Hospital Survey, which collects patients’ perspectives of hospital care, impacts hospital reimbursement from the Centers for Medicare and Medicaid.¹¹

Despite a growing interest in patient engagement, little work has been conducted to study patient engagement in the inpatient setting.^12–15 As noted in a recent New England Journal of Medicine article, access to information by patients is becoming more common, but providing access in the inpatient setting is particularly complex.¹⁶ To our knowledge, no one has studied the impacts of providing patients in the hospital with full access to their medical record. This includes data like medications, laboratory results, radiology and pathology reports, clinical summaries, operative/procedure reports and progress notes. We wanted to understand the clinician perspective on increased patient access to information as well as study the patient experience in receiving greater access to their information while in the hospital. Our hope was that by providing patients with access to their full clinical chart, they would feel more engaged in their care.

Methods

This research consisted of two parts; a survey administered to clinicians as well as a complementary field study conducted with patients in the hospital. In the field study, patients received daily printouts of their medical chart and then participated in semi-structured interviews to discuss their experiences. The study was conducted at Columbia University Medical Center (CUMC), a large urban academic center that is part of NewYork-Presbyterian Hospital. This research was approved by the medical center’s human subjects institutional review board.

Clinician Survey

A one-page survey was distributed to attendees of a staff meeting in a cardiac step-down unit, mostly comprised of nursing staff and physician assistants. It was also distributed at a monthly meeting of CUMC’s Housestaff Quality Council, a group of resident physicians representing each clinical department at the medical center. Additional surveys were collected from clinical staff on the cardiology floor.

The survey consisted of two sections, one that focused on clinicians’ perceptions of sharing different types of information with patients (e.g. medication information, clinical notes), and a second section that asked about potential consequences of increased patient access to information. Each of these items was evaluated on a 5-point Likert-type scale from ‘Strongly Disagree’ to ‘Strongly Agree’. Questions used negatively and positively worded stems to guard against acquiescence.¹⁷ The survey also provided additional space for participants to provide comments and encouraged them to elaborate on potential concerns, burdens, or benefits of increased patient access to information.

Responses to the survey were tabulated and analyzing using both Microsoft Excel 2010 and the R Statistical package.¹⁸ A descriptive summary of the data consisted of calculating frequencies, medians, and inter-quartile ranges (IQRs). Kruskal-Wallis analysis was completed, and a Bonferroni correction used for post-hoc analysis.¹⁹

Inpatient Field Study

We provided cardiology patients with a daily printout of their hospital care, printed from the institution’s electronic health record (EHR).²⁰ After four consecutive days of receiving this printout, we conducted in-situ semi-structured interviews with patients to elicit their feedback. Interviews were audio recorded, and transcripts of the audio were thematically coded.

Medically stable patients on a cardiology floor were approached for potential participation based on their ability to speak English and their anticipated date of discharge. In line with our institutional policies, their attending physician described the study and referred interested patients to the primary researcher (JP), who obtained informed consent.

After providing consent, the patient received a daily printout of his/her medical record, which included new information added to the EHR in the previous 24 hours. Specifically, the printout contained: laboratory test results, physician progress and consult notes, radiology reports, pathology reports, cardiology test results, discharge planning materials, operative reports, nutrition notes, the medication administration record, and other nursing documentation.

Minor editing was performed to remove phone numbers and social security numbers, and to delete unnecessary white space. No additional changes were made to the documents. A cover page describing the potential contents of the printout was added (Figure 1). The report was printed and hand-delivered each morning to each of the participants. The reports ranged from 5–40 pages each; 10–15 pages was typical.

Figure 1.

Cover page for patient printout.

After receiving daily reports for four days, a semi-structured interview was conducted. In addition to basic demographic questions, patients were asked to describe various experiences related to receiving their clinical information. Topics that were discussed included:

• Patients’ use of the printout (e.g., frequency, time spent)

• Patients’ perceived capacity to understand the information

• Preferences regarding structure/format of the information

• Patients’ desire to receive such information in the future

• Behavior changes as a result of having the information (e.g., asking more questions)

Interviews were audio-recorded and transcribed. As done in previous studies by this research team,¹³ transcripts were then thematically coded by the primary researcher (JP).²¹

Results

Clinician Survey

Approximately 65 surveys were distributed and 53 clinicians responded: 21 physicians (40%), 20 nurses (38%), 7 physician assistants (13%), and 5 allied health professionals (e.g., nutrition specialists, social workers) (9%). Responses are shown in Table 1 along with frequency, median, and IQRs in ordinal number format (1=Strongly Disagree; 5=Strongly Agree). Only the first question (sharing some information in general) had an IQR equal to zero. Six questions had an IQR of two. The Kruskal-Wallis test analyses differences between the groups (Physicians vs. PAs vs. Nurses vs. Other). Only the question on note-writing behavior was statistically significantly different between groups (p=0.033); however, post-hoc pairwise analysis of between group differences with a Bonferroni correction did not find any significance. The survey results are also presented using diverging stacked-bar graphs in Figures 2–4.²²

Table 1.

Summary statistics for clinician survey on attitudes toward inpatient information sharing.

	Disagree	Strongly Disagree	Neutral	Agree	Strongly Agree	No Response	Range	Q1	Median	Q3	IQR	Kruskal-Wallis
I am comfortable with patients having access to:
Some information from their EHR (in general)	0 (0%)	0 (0%)	6 (13%)	28 (62%)	11 (24%)	9 (20%)	3–5	4	4	4	0	0.458
Medication Information	0 (0%)	0 (0%)	1 (2%)	23 (45%)	27 (53%)	3 (6%)	3–3	4	5	5	1	0.336
Care Team Profiles	3 (0%)	0 (6%)	8 (15%)	23 (44%)	18 (35%)	2 (4%)	2–5	4	4	5	1	0.419
Lab Results	2 (2%)	1 (4%)	3 (6%)	23 (44%)	23 (44%)	2 (4%)	1–5	4	4	5	1	0.511
Radiology Results	3 (2%)	1 (6%)	5 (10%)	21 (40%)	22 (42%)	2 (4%)	1–5	4	4	5	1	0.088
Pathology Results	2 (2%)	1 (4%)	3 (6%)	24 (46%)	22 (42%)	2 (4%)	1–5	4	4	5	1	0.211
Operative/Procedure Reports	1 (2%)	1 (2%)	6 (11%)	27 (51%)	18 (34%)	1 (2%)	1–5	4	4	5	1	0.345
Progress Notes	10 (6%)	3 (19%)	11 (21%)	19 (36%)	10 (19%)	1 (2%)	1–5	2	4	4	2	0.092
Consultation Notes	7 (4%)	2 (13%)	12 (23%)	20 (38%)	12 (23%)	1 (2%)	1–5	3	4	4	1	0.233
In my opinion, providing patients increased access to their medical information will result in:
Increased time at bedside	9 (4%)	2 (17%)	15 (29%)	20 (38%)	6 (12%)	2 (4%)	1–5	3	3	4	1	0.970
Increased patient misunderstanding of information	15 (2%)	1 (28%)	11 (21%)	20 (38%)	6 (11%)	1 (2%)	1–5	2	3	4	2	0.541
Increased patient anxiety	14 (10%)	5 (27%)	13 (25%)	14 (27%)	6 (12%)	2 (4%)	1–5	2	3	4	2	0.638
Increased legal liability	11 (2%)	1 (21%)	16 (31%)	16 (31%)	8 (15%)	2 (4%)	1–5	3	3	4	1	0.784
More work for me as a caregiver	16 (0%)	0 (31%)	14 (27%)	18 (35%)	4 (8%)	2 (4%)	2–5	2	3	4	2	0.333
Increase in patient questions	5 (2%)	1 (10%)	17 (33%)	25 (49%)	3 (6%)	3 (6%)	1–5	3	4	4	1	0.198
Improved health behaviors	16 (4%)	2 (31%)	6 (12%)	26 (50%)	2 (4%)	2 (4%)	1–5	2	4	4	2	0.140
Enhanced patient-clinician communication	1 (0%)	0 (2%)	18 (35%)	30 (58%)	3 (6%)	2 (4%)	2–5	3	4	4	1	0.301
Increased patient engagement in decision making	2 (0%)	0 (4%)	16 (31%)	28 (54%)	6 (12%)	2 (4%)	2–5	3	4	4	1	0.570
I would change the way I write my clinical notes if I knew a patient could view them:	15 (4%)	2 (31%)	12 (25%)	15 (31%)	4 (8%)	6 (13%)	1–5	2	3	4	2	0.033*

Figure 2.

Clinician perspectives on patient information sharing.

Figure 4.

Clinician perspectives on expected changes in note-writing with increased patient information sharing.

Inpatient Field Study

Eight patients participated in the field study (Table 2). Seven completed the follow-up interview after receiving printouts of their hospital record for four days. The eighth patient was discharged prior to completing four days of receiving his records.

Table 2.

Participant Demographics of Inpatient Field Study.

	Sex	Age	Admission Reason	Presenting Chief Complaint	LOS*	Highest Level of Education	Reports Using Internet for Health Information
P1	Female	87	Congestive heart failure exacerbation	Shortness of breath	6 days	High School	No
P2	Male	56	Diarrhea, atrial fibrillation	Diarrhea, shortness of breath	22 days	Some College	Yes
P3	Male	65	Septic Shock	N/A†	8 days	Graduate School	Yes
P4	Male	43	Myocardial infarction	Chest pain	4.5 months	College Graduate	No
P5	Male	35	Hemoptysis	Hemoptysis, shortness of breath	5 days	College Graduate	Yes
P6	Male	54	Awaiting Heart Transplant	N/A‡	5 weeks	Associates Degree	Yes
P7	Male	54	Ventricular tachycardia	Shortness of breath	5 weeks	Some College	No

^*Length of stay at time of interview

^†Transferred because of fevers from inpatient acute rehab

^‡Admitted for heart transplant listing

Overall, patients’ feedback on receiving the daily printouts was positive. Out of the seven patients we interviewed, six patients indicated that they would want to receive this type of information again if they had to return to the hospital. All six expressed a desire to continue receiving the printouts for the remainder of their current stay. Key themes that emerged from the semi-structured interviews were: variations in use, difficulty understanding medical jargon, appreciation of having the information, suggestions for improvement, changes in interaction with clinicians, and changes in engagement.

Variation in Use

There was a range in use of the printout. Two patients stated that they did not look at the printout (P5 and P7), P4 looked at it on two of the four days, and the remaining patients reported looking at it every day. The scope of use also varied. Some patients reported only skimming the information, while others read it in detail and even added their own annotations. For example, P2 commented: “I take a look, and if I see something that’s not [in range], I just mark it.”

Difficulty Understanding Medical Jargon

The printouts were created from the EHR without any substantive changes made to the content or format. Patients discussed having particular difficulty with certain medical terms and acronyms. P3 stated, “I read until I found technical stuff, and then I would jump over it…the parts I read, and I understood, were interesting and beneficial.” P2 said, “Most of the things I understand, but some of the other things…like letters in combination, ‘WT’ or whatever… [were] not of use to me.”

Several patients reported looking for specific items which they knew about. P1 specifically asked about her creatinine values and searched for the results of a sonogram performed the previous day. P5 expressed difficulty with understanding some material, but was planning to use the Internet to supplement his knowledge: “I’m familiar with some stuff,” he said, “but…some of the terms I’m not familiar with, so once I have my laptop I’m probably going to take another look at it.”

Appreciation in Having the Information

Patients expressed appreciation for having access to information about their care that they wouldn’t normally receive. P7, who did not review the printouts at all, reported that it was important to him to “have a file.” P2 said it was “very, very neat…because they give you all the details,” and P5 stated: “Notes help a lot too, to see if the doctors are interpreting what I say… and they’re pretty right on. And just [to see] what they thought as far as the possible diagnosis.”

Suggestions for Improvement

In addition to wanting more explanation around medical terms and acronyms, patients expressed ideas regarding changes to the process that would make the intervention more helpful. For instance, P5 requested that there be a more clear delineation of the author of notes: “Sometimes I see a bunch of comments, and I don’t see exactly which doctor said it.” P3 expressed the desire for more descriptive information: “If you could give me more information describing my condition and my situation…more than the technical stuff…more information as far as what the doctors are finding, and what the doctors are doing.”

We also discussed the possibility of presenting the information to patients via a computerized interface. Most patients (four of the five who viewed the printouts) expressed an interest in viewing their information on a tablet computer or laptop.

Changes in Interaction with Clinicians

We asked patients if having access to their information prompted them to ask follow-up questions. Only one of the seven patients (P5) mentioned asking anything related to information in his printout. Several patients acknowledged the busy clinical workload of the providers and did not want to further impose on their limited time. P2 said, “[My doctors] have one or two minutes, [then] they’re running all the time.”

Changes in Engagement

Patients expressed feeling more informed about their health as a result of receiving the daily printouts. P2 mentioned that “if you don’t ask, you don’t receive [information], but with that paper, I am very on top of the situation.” P5 said, “Instead of just listening to the doctor…I should pay attention to myself too” and believed the printouts could help him to do so.

In addition to feeling more informed, multiple patients mentioned the ability to act as a ‘fact checker’ of the information. P1 reported finding a prior procedure that was recorded with an incorrect date. P5 said it was, “neat to see the notes because sometimes during the interviews with the doctors, they’ll write things down that I said, and well I didn’t really mean it exactly that way.” P5 also said, “I see the hospital medications…that’s important I guess, just in case there are any discrepancies…between what stuff I think I should be taking, and [that] which they didn’t give me here, and I’m like why, not?” P2 and P6 mentioned noticing discrepancies between when medications were documented as administered and when they believed them to have actually been administered.

Discussion

While patient engagement has recently been a topic of considerable interest^23–29, research in the inpatient setting is lacking.^12,14,15 In The Patient Checklist, Elizabeth Bailey observed that “once a patient enters a hospital for treatment of any kind, what he or she needs most of all is knowledge – what is happening to him, and why.”³⁰ We believe our study may be the first to evaluate the effect of providing patients with access to their entire charts, including physician notes, during their hospital stay. The most important findings of our study were that information sharing: 1) was perceived as desirable to patients and acceptable to clinicians, 2) allowed patients to more actively participate in their hospital care, and 3) may impact clinician behavior in terms of workload, communication, and note-writing practices.

Our results suggest that greater information sharing with hospital patients can be beneficial. Of the seven patients we interviewed, six of them requested to continue receiving the daily printouts. Patients appreciated seeing the details of their hospital care even in the raw, unfiltered format that came from our EHR system. We believe that a more tailored format, though difficult to actualize, could have even greater potential to increase patient engagement. The survey results indicated that clinicians also viewed information sharing with patients favorably. For clinicians, the sharing of objective data in particular seemed to provoke little controversy. The sharing of more subjective data, such as progress notes and consultation notes, was less agreeable to survey respondents, but still, the majority were comfortable with sharing this information.

Clinicians’ expectations of the consequences of inpatient information sharing were mixed. Most were optimistic that sharing information would enhance communication and increase patient engagement in decision-making. However, some survey responses reflected concerns about increased anxiety and misunderstanding of information. Though our sample was small, none of the patients interviewed expressed that they felt additional anxiety from viewing their records. On the contrary, several patients commented that receiving the daily printout made them feel more informed and better able to understand the details of their care. Additionally, having access to their information allows patients to act as another ‘line of defense’ to identify incorrect information and potentially decrease medical errors.

In the outpatient setting, the OpenNotes project explored the effects of giving patients the ability to read their doctor’s office visit notes.³¹ The project has been successful enough that the three institutions involved continued sharing notes with patients after the study ended. Additionally, the project has expanded to other institutions, including the Department of Veterans Affairs (VA).³² Although further study of information sharing in the hospital setting is warranted, our results suggest that the OpenNotes concept may also be applicable to inpatient care.

OpenNotes investigators reported that increased access to physician notes by patients may change note-writing behavior.³³ Comments from our clinician survey suggested that this was a concern at our institution. For example, patients may be offended by reading certain descriptions in their charts (e.g., “obese”, “disheveled”). Moreover, notes often contain expressions of uncertainty, frightening differential diagnoses, and findings that lack interpretation. If notes are readily accessible to patients, clinicians might avoid using certain terms, or omit clinically relevant details to protect patients from unnecessary anxiety. On the other hand, because so much of what clinicians currently document is not easily understandable by patients, they may be inclined to write in a manner that is more intelligible to patients (e.g., by using fewer acronyms).

Design Considerations for Inpatient Engagement

Beyond a pilot study, printing and hand-delivering daily reports as we did in this study is probably not a feasible option for most hospitals. However, a recent study from a neonatal intensive care unit described delivering a one-page paper handout to parents each day of their baby’s stay.³⁴ The handout included information on the baby’s care team, respiratory status, nutritional status, medications, most recent lab results, and care plan. Similarly, the VA began providing inpatients with access to limited information through the Daily Plan project, which delivered a daily printout of the patient’s medications, appointments and diagnostic tests which nurses would then review with the patients each day.³⁵ We¹³ and others^14,15 are working to deploy online patient engagement solutions for hospital patients. So far, these projects have provided tablet computers to patients in the hospital on a small scale and with limited information—excluding clinical notes, for example.

Future work should investigate how to present patient-specific medical information in a more patient-friendly manner. Appealing to diverse populations of patients, including those with low health literacy, low literacy in general, and non-native English speakers is a particular challenge. Using resources such as the Consumer Health Vocabulary and MedlinePlus may help to make complex medical terminology more understandable.^36,37 Additionally, use of visualizations has been shown to help patients better understand health information, especially for individuals with low health literacy.^38–40 Research on the creation of visualizations that explain inpatient data is warranted.

Another challenge to providing increased access to patient information is addressing privacy and security concerns. A recent article by Bates and colleagues discusses the need for family, caregiver, and care partner access to patient data.⁴¹ Restricting access to certain types of data may be a feature necessary for privacy of sensitive topics like testing for drug use, or pregnancy in teenagers.

Limitations

This study was limited by a relatively small sample size for the clinician survey (n=53), and similarly by a small sample size of hospital patients who participated in the inpatient field study (n=8). Of the eight patients in the inpatient field study, one was discharged prior to the interview and two additional patients had not read through their records. Thus, the sample size was insufficient to achieve thematic saturation. Nevertheless, we believe our results shed light on the understudied topic of engaging hospital patients in their care and provide a foundation for future research. The study was conducted at a single site, an academic medical center in an urban setting, and our findings may not generalize to other settings. More specifically, our study was focused primarily in the domain of cardiology, and different results may be found in other patient populations. For example, patients with cardiac disease may have higher baseline engagement levels than patients with less chronic diseases, and thus may be more interested in seeing their data. Future work should explore whether differences in settings or patient populations exist with respect to inpatient engagement.

Conclusion

Healthcare delivery organizations are moving towards greater sharing of information with patients. Consumers, who are the “sole subject matter expert on themselves,”⁴² are increasingly expecting to have access to their data. Our study found that clinicians were mostly comfortable with increased information sharing, and patients benefitted from receiving daily printouts of their hospital care record. Increased access to information will enable patients to more actively participate in their hospital care.

Figure 3.

Clinician perspectives on consequences of increased patient information sharing.