Before any invasive procedure, physicians have a legal obligation to inform patients. Traditionally, this involves a discussion with a physician, supplemented by written leaflet information directed at the specific procedure.
Comparison of the use and effectiveness of computer-based visualization opposed to standardized conversation for providing patients with information of forthcoming procedures (coronary catheters or endoscopy procedures).
Prospective, randomized trial with 56 participants allocated in two different groups: Visualization Group (standardized information supported by a tool for displaying two-dimensional pictures to explain medical facts as well as informative leaflet) or Control Group (standardized information and informative leaflet only). Detailed information was given about the indication, the probable complications and the details of the forthcoming procedures (coronary catheters or endoscopy procedures). All participants had to reach a Karnofsky Score of 70 points and be able to understand German or English. Main outcome measures were patient's satisfaction with physician-patient conversation, patient's acquired knowledge and duration of the intervention as described above.
Patients of the Visualization Group were more satisfied with the conversation and had higher knowledge scores after the conversation. A Mann-Whitney-U-Test between the two groups showed that these differences in satisfaction (P<0.001) and knowledge (P=<0.006) were statistically significant. Length of time needed for the conversation was slightly higher in the Visualization Group, but this difference was not statistically significant (25 versus 23 min;
Using computerized visualization increased the satisfaction and knowledge of the patients. The presentation of the visualized information in the Visualization Group did not demand significantly more time than the standard conversation in the Control Group.
During the last decade we have observed an increasing demand for better integration of patients in clinical and ambulatory health care [
The information has to present the available evidence in a form that is acceptable and useful [
Since the beginnings of human communication, learning and comprehension have always been supported graphically. In education, pictures often clarify difficult facts better than written language. In anatomy, for example, drawings by Netter explain the human body [
Educational materials designed to deliver information and support a more active participation of patients in health care decisions can be effective tools for empowering patients [
Most multimedia tools and information brochures serve as a source of information for patients lacking a professional adviser. It has been reported that in too many cases the information contained in patient information leaflets is inaccurate or misleading [
Mobile computer at patient bedside
Dr Topf's patient information system makes use of the graphical presentation of medical content during the conversation between the physician and the patient to give the patient a quick and extensive understanding of the medical facts [
Screenshot Dr Topf
Participants included 56 patients of a cardiology ward and a gastroenterology ward. The patients were examined over a period of 5 months (
The sample size calculation was determined by the measured effects of our pilot study. As a result of the pilot study, satisfaction and knowledge of the patient obtained effect sizes between 0.65 and 0.71. For a parametric test comparing two independent groups with an assumed power of 0.8 and a level of significance (alpha = 0.05), 26 patients were perceived as an ideal number for each study group [
For the allocation of the patients to one of the study groups, every physician received eight sealed envelopes. The inscription on the envelopes only indicated the name of the physician and the kind of procedure (cardiology or gastroenterology). Four of the envelopes contained method A (standardized information supported by computer-based picture material), the other four contained method B (standardized conversation). The proportion of four patients per intervention group was equally divided into cardiology and gastroenterology procedures. To prevent the case of double information, all physicians were told that one of their envelopes had been given to another physician. This implied that the ratio of method A to method B could have changed from 4:4 to 5:3 or 3:5. As a consequence the physician would remain blinded from his first to his last patient. In the course of this study no change of the ratio was needed, so a balance of 4:4 for each physician was guaranteed.
Flow diagram of patients through trial
After signing a written declaration of consent, patients were randomly assigned to one of two groups via the random envelopes as previously described . The intervention group received standardized information supported by picture material (i.e. a sample of five pictures maximum was presented on a sub-notebook at patient bedside). The computerized presentation was limited to 5 minutes. Physicians who were taking part in the study had been trained to handle the information tool before the trial began.
A second group was informed by means of standardized conversation by a physician. This group was referred to as the "Control Group" because this procedure is the most common way of informing patient in Germany. Participants of both groups received the same informative brochure [
Seven physicians (4 senior house officers, 2 residents, 1 junior house officer) had to inform four patients of each group. Before providing information to the patient, they had to report to the study supervisor whether a patient met the criteria of inclusion.
The physicians gave every participant detailed information about the indication, the probable complications, and the details of the forthcoming procedures (i.e. about anatomy, pathology, complication ratio, possible side effects, postinterventional behavior, and alternative interventions). The following procedures were taken into consideration:
Right-cardiac catheter
Left-cardiac catheter and coronary catheter
Percutaneous transluminal coronary angioplasty (PTCA)
Electrophysiologic catheter of the right heart
Endoscopic retrograde cholangiopancreatography (ERCP)
Gastroscopy
Colonoscopy
A list with all necessary contents regarding each procedure was given to the physicians.
Consequently, they had to give every participant detailed information about the purpose of the procedure (pathological changes such as ulcers, varicose veins, sources of bleeding, polyps, or tumors), alternative ways to the procedure (e.g., x-ray, surgery), the probable complications and their treatment (e.g., punctured or injured colon wall requiring immediate surgery; bleeding, which can be treated by injection of drugs; allergic reactions), and the appropriate postsedation behavior (bed rest, no food or liquids for at least 1 hour after the examination).
The procedure was carried out one day after providing the information to the patient.
After the intervention, every physician completed an anonymous numbered protocol to determine the time spent on the conversation, the time used for visualization, the method of intervention, the kind of procedure, and any important questions asked by the patient.
Shortly after the conversation the patient was asked to personally assess the quality of the physician-patient conversation via a patient satisfaction questionnaire (
The evidence of the visualized approach was evaluated using a formalized questionnaire (standard of knowledge). Ten multiple choice questions taken from assessment papers for medical students and adapted to patient knowledge level were used to assess the method of patient education. For every query the patient had to choose either a correct or a wrong statement of five probable statements. A total score of 10 points could be reached. Higher scores indicate greater knowledge (
Patient satisfaction questionnaire
Scale reliability was calculated for patient's satisfaction as internal consistency (Cronbach's alpha coefficient) for the total sample population. Baseline data was collected before randomization. To check whether the assessment criteria correlated with the patient's educational level, age, and the time allocated to the conversation, a Mann-Whitney U test or
The Mann-Whitney U test was used for the comparison of patients' satisfaction and knowledge in both groups.
Allocations were sealed in opaque numbered envelopes that were opened by the physician after instruction by the independent study supervisor who generated the allocation sequence. Questionnaires had been handed out to the patients by an independent observer who was not informed about which group each patient was in. The statisticians had no contact with study participants and received only unblinded data.
Between June and October 2002, a total of 62 patients were identified as potential participants. Of the remaining 60 patients who met the criteria of inclusion, 56 received the allocated intervention (see
The length of time needed for the conversation was analyzed for 86% interventions (see
As the performed
The study subjects ranged in age from 22 to 91 years. The average age was approximately 57.5 (SD 13.8) years (
Sociodemographic data of the Visualization- and Control Group
Variable | Visualization Group |
Control Group |
|
|
Age (years) | 0.498 | |||
Mean age ± SD |
55.7 ± 10,35 | 58.2 ± 11.6 | ||
Gender | 0.717 | |||
Female | 11 | 8 | ||
Male | 17 | 20 | ||
Professional qualification | 0.666 | |||
N.A. |
3 | 1 | ||
Apprenticeship | 9 | 18 | ||
Craftsman/technical school | 6 | 2 | ||
Technical college/university | 3 | 3 | ||
No graduation | 7 | 4 |
* The group differences were calculated using
† SD = standard deviation; N.A. = not announced
Outcome measures of the Visualization- and Control Group
Visualization Group |
Control Group |
||||||
Variable |
|
|
|
|
|
|
|
Patient satisfaction questionnaire |
25 | 21.2 (19.2 to 23.8) | 4,8 | 28 | 15.8 (14.1 to 17.5) | 4,5 | <0.001 |
Item no. 5 | 4.1 (3.50 to 4.69) | 1.44 | 2.9 (2.46 to 3.34) | 1.13 | |||
Knowledge questionnaire |
24 | 7.21 (6.5 to 7.9) | 1,6 | 25 | 5.04 (3.3 to 6.2) | 2,8 | 0.006 |
Knowledge questionnaire |
24 | 7.21 (6.5 to 7.9) | 1,6 | 25 | 5.04 (3.3 to 6.2) | 2,8 | 0.006 |
Knowledge questionnaire |
24 | 7.21 (6.5 to 7.9) | 1,6 | 25 | 5.04 (3.3 to 6.2) | 2,8 | 0.006 |
Overall time |
25 | 10.16 (8.55 to 11.24) | 3,0 | 23 | 9.23 (7.19 to 11.28) | 4,8 | 0.441 |
Time for visualization |
3.54 (3.41 to 4.40) | 1.2 |
* The group differences were calculated using
‡ possible range 5 - 25 points
† possible range 0 - 10 points
§ M = mean score; CI = 95% confidence interval, SD = standard deviation
Cronbach's alpha coefficient for the internal consistency of the patient satisfaction questionnaire was 0.94 and can be considered good (
In the total knowledge score, the patients of the Visualization Group reached 2.2 points more than the patients of the Control Group (95%-CI = [0.9 to 3.43]) (see
No major differences were seen between the length of time needed for the conversation of the analyzed 48 patients of the Visualization and Control Groups (average time,
"I do not feel that a presentation of the images takes up significantly more of my time. However, as a consequence, the patient wants to learn more about his disease from the physician."
"The laptop computer did not attract the attention of the patient too much. I had the impression that the patient quickly picked up the physiological-pathological information and was able to ask further specific questions."
"Letting the physician operate the program seems more effective to me than having the patient look at such images by himself."
In this prospective, randomized trial, we hypothesized that computer-based visualization would support a conversation for providing patients with information about forthcoming procedures. The patient's satisfaction with the conversation revealed higher satisfaction scores. In spite of the high reliability score of the internal consistency (0.94), sufficient variance in the scale of the patient satisfaction questionnaire was found. As a main focus, the impact of the computer-based visualization tool was directly addressed by our questionnaire, which showed a difference of 1.2 points. This means a difference from "it applies" (3 points) to "it applies very well" (4 points).This observation is consistent with other reported results [
The time needed for the conversation between physician and patient when supported by visualization was one of the most important points of interest. Some physicians pointed out that the supplement of visualization did not take more time compared to the standardized conversation (see
While in the present study the software ran on a laptop computer and was brought to the patients' bedside, the information, which is implementated in HTML, could alternatively be distributed to patients via the internet prior to hospitalization. In the future, patients undergoing elective procedures could be empowered at home or in the general practitioner's office before hospitalization. In this study our main focus was on the examination of patient empowerment by physicians assisted with computer-based visualization for already hospitalized patients.
By the increase of knowledge in the Visualization Group, it could be assumed that visualization effectively supports the educational process. Although other studies have evaluated patient satisfaction with computer-assisted instruction, few have evaluated patient knowledge before forthcoming procedures [
Another concern frequently voiced by physicians during the pilot phase of testing was that the visualization could raise patients' anxiety. In this study none of the patients mentioned or expressed concerns in any other way that would support this hypothesis.
After our pilot study we decided to continue the study with several physicians and one independent observer to minimize the Hawthorne-Effect [
Our findings show that computer-based visualizations like "Dr Topf´s patient information system" have desirable effects on the patient's satisfaction and knowledge. Research into improving health care by visualization of medical content should be intensified. Following the line of argumentation of Faden and Beauchamp [
We thank our colleagues at the Medical University Hospital, Heidelberg, and the patients whose contributions made this study possible. We had benefit from the development of "Dr Topf's Patient information system" and the program "The Countrywide Non Communicable Disease Intervention (CINDI) Program" by Günter Topf MD and Egbert Nüssel PhD from the WHO Collaborating Center.
Contributors: ME, the principal investigator, participated in the research design, collected the data, undertook the preliminary analysis of the results, and was primarily responsible for writing the paper. HBB identified the need for a trial and designed the trial; coordinated the project, guided the collection, analysis, and interpretation of data, and helped write the paper. NK participated in the design of the trial and contributed to the writing of the paper. BW undertook the data analysis and contributed to the interpretation of the results. KM advised on the design of the trial and assisted in the interpretation of findings. WH conducted the project. AH provided many helpful suggestions during study design development and preparation of the manuscript. ME will act as guarantor for the study.
None declared.