Fax: (011) 39 0372 405235
Effects of media information on cancer patients' opinions, feelings, decision-making process and physician-patient communication
Article first published online: 16 JAN 2004
Copyright © 2004 American Cancer Society
Volume 100, Issue 5, pages 1077–1084, 1 March 2004
How to Cite
Passalacqua, R., Caminiti, C., Salvagni, S., Barni, S., Beretta, G. D., Carlini, P., Contu, A., Di Costanzo, F., Toscano, L. and Campione, F. (2004), Effects of media information on cancer patients' opinions, feelings, decision-making process and physician-patient communication. Cancer, 100: 1077–1084. doi: 10.1002/cncr.20050
- Issue published online: 18 FEB 2004
- Article first published online: 16 JAN 2004
- Manuscript Revised: 3 DEC 2003
- Manuscript Accepted: 3 DEC 2003
- Manuscript Received: 30 JUN 2003
- Italian Ministry of Health
- cancer information;
- decision-making process;
- physician-patient relationship;
- mass media
The objective of the current study was to determine the influence of media information on the opinions and feelings of patients with cancer and to measure the factors that affected the decision-making process and physician-patient communication.
The study consisted of a sequence of 2 nationwide surveys across the same dynamic target population of 2600 unselected patients with cancer who attended 1 of 13 centers throughout Italy. The authors measured the changes in patients' opinions and attitudes at the peak of a media campaign promoting the Di Bella therapy, an unproven cancer treatment method, and after the publicized demonstration of its ineffectiveness. An identical 10-item questionnaire was used.
Opinions and feelings changed in the two surveys according to the way the media described the efficacy of the treatment, but physician-patient communication and the decision-making process remained unchanged. Multivariate analysis confirmed the enormous influence of the media on patient opinions (odds ratio [OR], 4.67; P < 0.0001), feelings of hope (OR, 3.63; P < 0.0001), and confusion (OR, 0.51; P < 0.0001), but not on physician-patient communication or the decision-making process. Educational level influenced almost all of the studied factors, and communication and decision-making also were influenced by the patients' gender and place of residence. There was no significant correlation with patient age.
The media play a powerful role in affecting patients' opinions and feelings; the physician-patient communication and the decision-making process are not subject to media influence but are related primarily to level of education. The power of the media should be directed toward improving the spread of scientific knowledge to encourage behavioral changes, particularly among individuals with lower levels of education. Cancer 2004;100:1077–84. © 2004 American Cancer Society.
For patients with cancer, the news provided by the media represents a major source of information on new therapies and new methods of cancer prevention. However, the information is sometimes incomplete or distorted and thus induces less-than-optimal behaviors among the patients or raises false optimism about the efficacy of therapies.1, 2 At the end of the 1990s, Italy experienced an unprecedented media campaign favoring an unproven alternative method of cancer treatment known as Di Bella therapy (DBT), which consisted of a cocktail of somatostatin, vitamins, melatonin, and other drugs that were mixed on an individual basis and that could cost up to $5000 (U.S.) per month. This unexpected and particularly intense campaign led to a national debate that divided public opinion and to a massive demand from oncologic patients that its administration be made available free of charge in the name of freedom of choice.3, 4 In January 1998, public pressure forced the Ministry of Health to organize a series of independent, Phase II protocols to test the efficacy of the therapy in various tumors, and finally, 6 months later, the results of these studies showed the ineffectiveness of DBT.5 The consequences of the ‘Di Bella event’ confirmed the need to investigate the ways in which health information is given, the factors that influence the physician-patient relationship, and the decision-making process that patients adopt when choosing a treatment.
To understand these aspects in more detail, we carried out two cross-sectional studies: the first at the time of the peak of the media campaign promoting the DBT and the second immediately after the release of the results of clinical trials demonstrating the inefficacy of this alternative therapy.5 Both studies were based on the use of the same questionnaire distributed to consecutive patients attending 13 Italian oncologic centers. The results of the first survey revealed 1) the enormous influence of the media on patient feelings and their opinion regarding the efficacy of this alternative treatment, 2) that the main reasons underlying their choice of treatment were their hopes regarding its efficacy and their trust in the physician recommending it, and 3) that serious inadequacies in physician-patient relationships largely contributed toward the choices and behavior of the patients during the course of this event.6
In the current report, we present the results of the second survey and analyze and discuss the overall results of the two surveys with the objectives of 1) describing how patient opinions, attitudes, and decisions change in relation to the way the media describe the efficacy of a given treatment; 2) identifying and measuring the weight of the factors affecting the feelings and attitudes of patients, physician-patient communication, and the decision-making process; and 3) providing indications useful for the introduction of strategies aimed at preventing or limiting the damage caused by campaigns that provide inaccurate information. The current large-scale, nationwide study demonstrated that the health information offered by the mass media profoundly influences the opinions and feelings of patients, but not physician-patient communication or the decision-making process adopted by patients in choosing treatments, which are determined by their sociodemographic characteristics and mainly by their level of education. We also offer a measurement of the weight of age, gender, education, and place of residence in determining patient attitudes and behaviors.
MATERIALS AND METHODS
The current cross-sectional study consisted of a sequence of two nationwide surveys that were conducted across the same dynamic target population: the first took place between February 25, 1998, and March 31, 1998, when public debate had reached its peak and just before the beginning of the Phase II protocols designed to test the efficacy of the therapy; the second was conducted between December 20, 1998, and January 31, 1999, after the results of the Phase II trials demonstrating the inefficacy of the DBT had been publicized by the Italian Ministry of Health and made known to the population by the media.
The objective of the study was to investigate overall opinions and attitudes in large populations and to determine how these change with variations in media information. For this reason, all of the patients attending the oncology centers on the established days were surveyed, regardless of the length of time that had elapsed since their diagnosis of cancer or the reason for their visit. The 2 surveys were carried out in the same oncology centers using a previously described, 10-item questionnaire, which is illustrated in Table 1.6 The surveys involved 13 medical oncology centers in various parts of Italy; the questionnaire was distributed consecutively over a period of 2–5 days to the first 100 unselected patients attending each center for adjuvant or palliative chemotherapy or for a follow-up visit. All patients age ≥ 18 years who had been diagnosed with a malignant neoplasm and who did not suffer from any cognitive or physical impairment that would prevent them from completing the questionnaire were considered eligible for the study. The study was designed not to follow individuals but to follow the population; therefore, no information was collected regarding the type of oncologic treatment the patients were receiving or whether they had participated in the first survey. A nurse distributed the questionnaire, and the patients returned it using an ad hoc collection box.
|Please cross only one reply, even when there are more than two alternatives.|
|1) Do you know something about the “Di Bella treatment”?|
|2) What was the main source of your information?|
|Daily, weekly press|
|Relatives, friends, acquaintances|
|The testimonies of other patients|
|My general practitioner and/or specialist|
|Directly from Dr. Di Bella or one of his collaborators|
|3) On the basis of the information you have received, do you think the “Di Bella treatment” is:|
|4) Since you became aware of the “Di Bella treatment,” has your hope of being cured:|
|5) Do you communicate with your doctor/oncologist about new treatments?|
|I would like to, but I can't|
|6) Since you heard that there may be another treatment such as that of Di Bella, have you felt:|
|7) In order to be cured, would you:|
|Only try tested treatments that give no absolute certainty of a cure|
|Also try unproven treatments that give some hope of a cure|
|8) When choosing a treatment, do you think it is better to:|
|Follow the findings of scientific research|
|Follow the testimonies of other patients|
|Follow the recommendations of a trusted doctor|
|Follow your own personal intuition|
|9) What kind of doctor would you trust most?|
|A doctor who does not suggest a new treatment until its efficacy has been proven|
|A doctor who suggests that you also try unproven treatments that may increase your chances of being cured|
|10) Would you personally try the “Di Bella treatment”?|
|Only if the other treatments no longer worked|
|Only if I were sure of its efficacy in my own case|
|Only if it were free of charge|
|Only if I were sure it would not harm me|
|Only if my own doctor prescribed it for me|
|In any case|
|In no case|
The questionnaire was completed anonymously and included a cover letter and a section asking the patient to provide sociodemographic information. The design and contents of the questionnaire were reviewed by a pool of expert psychologists. Before the beginning of the study, the questionnaire was administered to 30 patients attending the coordinating center in Parma to verify that the questions were comprehensible and that the patients could answer them. The cover letter explained the aims of the study and assured patient anonymity, confidentiality, and freedom to participate or not participate. The first section asked for general information (such as gender, age, and level of education), and the second section assessed the primary study objectives. The study received informal approval from the Italian Ministry of Health (Rome, Italy), the Italian Superior Health Institute (Rome, Italy), and the University Hospital of Parma (Parma, Italy).
All questionnaires were returned to the coordinating center in Parma, where the answers were validated and analyzed. The frequency of the responses to individual questions varied, because any question with more than one answer was considered invalid, and because some questions were not answered at all. To exclude selection bias, we compared the age, gender, education level, and geographic distribution (Northern, Central, and Southern Italy) of our sample with those of the Italian population reported in the 1991 Census.7
In this article, the word variables refers to the sociodemographic variables age, gender, level of education, and geographic residence, which were categorized as follows: age (< 50 years, 50–59 years, 60–69 years, or ≥ 70 years); level of education (primary school, ≤ 13 years; secondary school, 14–18 years; or completion of university schooling); and residence location (northern, central, or southern Italy). The chi-square statistic was used to test the differences in the percentage of responses to the questions between the two phases of the study.
To ascertain the factors that were important in determining the responses given in the two study phases, we used multiple logistic regression analysis, with the individual odds ratios (ORs) ± 95% confidence intervals for each factor estimated after adjusting for the other variables. For multiple-choice (nondichotomous) questions, the response of greatest interest was classified as 1 (if present), and all other responses were classified as 0. In addition to the sociodemographic covariates of gender, age, geographic origin, and education level (dummy variable), the role of the media in conditioning the responses was evaluated by considering the media information covariate as 1 in the first survey and 0 in the second survey. Age initially was investigated as a continuous variable in the multivariate model; however, because it did not prove to be of considerable importance, it was fitted to a new model that did not contain age. Comparison of the two models showed that the variable age added nothing to the model; therefore, we verified the possibility of replacing the age variable with three design variables based on age quartiles to determine whether these designs yielded a quadratic or U-shaped function. The estimated OR revealed no significant trend; thus, we chose not to include age in the final multivariate analysis.
The goodness-of-fit of the model was assessed by means of the Hosmer–Lemeshow test.8 The data were analyzed using the SAS System (version 6.11; SAS Inc., Cary, NC).
Characteristics of the Respondents
A total of 2600 questionnaires were distributed: 1300 for each survey. In the first survey, 1120 questionnaires (86.1%) were returned and analyzed, and 20 questionnaires (1.8%) were excluded because they contained no answers or were illegible. Of the 11,200 questions that were considered for analysis, 288 questions (2.6%) were excluded, either because more than 1 response had been given to a multiple-choice question or because a question had not been answered at all. In the second survey, 1042 questionnaires (80.1%) were returned and analyzed, and 25 questionnaires (2.4%) were excluded because they contained no answers or were illegible. Of the 10420 questions analyzed, 235 answers (2.2%) were excluded because > 1 answer had been given to a multiple-choice question or because no answer had been given. The patients who returned the questionnaires of the two surveys were similar in terms of age, geographic origin, and level of education (Table 2).
|Characteristic||First survey||Second survey||1991 Italian census|
|Geographic origina (%)|
Table 3 summarizes and compares the results of the two phases of the investigation. There was a statistically significant difference in the answers to all questions with the exception of the first questions (information about DBT). The patients declared that their main source of information was the television and/or radio (62% and 70% in the first and second surveys, respectively), followed by the press (26% and 17%), and only 5% declared that they were informed by their physician. In relation to the therapy, there were considerable differences in patient opinions between the first and the second surveys: a considerable decrease in the percentage of patients who considered it efficacious (from 42% to 14%) and increases in the percentage of patients who declared themselves uncertain (from 57% to 75%) and the number of patients who judged the therapy to be ineffective (from 1% to 11%).The second survey also revealed a decrease in the number of patients with hopes of being cured (Question 4; from 53% to 25%) and increases in the number of confused patients (Question 6; from 48% to 65%) and the number of patients who had a feeling of insecurity (from 5% to 11%).
|Questions||First survey (n = 1120) (%)||Second survey (n = 1042) (%)||P valuea|
|1) Knowing about DBT||0.88|
|2) Source of information||<0.01|
|3) Opinion of DBT||<0.0001|
|4) Feelings of hope after DBT||<0.0001|
|5) Communications with oncologist||<0.001|
|I would like to but I cannot||20||16|
|6) Feelings of security after hearing about DBT||<0.0001|
|7) Preferences for choosing a treatment||<0.0001|
|Proven without guarantee of cure||37||52|
|Unproven but with hope of cure||63||48|
|8) Source of advice on choosing a treatment||<0.01|
|Findings of scientific research||32||34|
|Other patients' advice||6||3|
|9) Reasons for trust in a physician||<0.01|
|Suggesting only proven treatments||53||60|
|Also suggesting unproven treatments||47||40|
|10) Patient willingness to try the DBT||<0.0001|
|In any case||11||3|
|Only under certain conditions||87||90|
|In no case||2||7|
We investigated the decision-making process by means of Questions 7–9, which particularly reflected the media debate of the time. One of these items (Question 7; the importance of hope in the choice of treatment) referred to a rather hypothetic situation, and the others referred to more concrete situations (the role of scientific research findings and attending physicians in determining choices and the factors influencing the patients' trust in their attending physicians). Hope in the efficacy of a treatment greatly affected therapeutic choices, although the second survey revealed a decrease in the number of patients who declared that they would prefer unproved therapies that offered a hope of cure (from 63% to 48%; Question 7). Likewise, it revealed an increase in the number of patients trusting physicians who only suggested therapies of proven efficacy (from 53% to 60%; Question 9). With regard to the sources of advice for choosing a treatment (Question 8), > 50% of patients declared that the main source was the advice of a trusted physician followed by the findings of scientific research, which was preferred by > 30% of patients. These findings were almost unchanged in the two surveys.
Also in terms of physician-patient communication (Question 5), the responses in both surveys mainly were unchanged; there was a slight difference in the number of patients who declared that they communicated with their oncologists (from 56% to 52%); however, about 50% did not have good communications with their oncologists.
The results of the logistic regression analysis (Table 4) make it possible to estimate the relative weight of each of the considered factors (the media campaign, education level, geographic origin, gender, and age) in determining patient responses. Age had no statistically significant effect on any of the questions and, thus, was not considered in the final model. For each question, the table lists the ORs relating to all of the analyzed covariates as a measure of their association with the answers of the patients: the stronger the statistical relation, the greater the effect of the given factor on the response.
|Question||Adjusted OR, 95% CI, and P values for controlled factors|
|Media information||Education level||Geographic origin||Gender||Hosmer–Lemeshow testb|
|Second survey||First survey||University||Secondary school||Primary school||Southern or central Italy||Northern Italy||Female||Male|
|Opinion of DBT (effective vs. other)|
|Feelings of hope after DBT (increased vs. other)|
|Feelings of confusion after DBT (increased vs. other)|
|Communications with attending oncologist (yes vs. other)|
|Decision-making process (follow physician's advice vs. other)|
Media information was the factor that had the greatest effect on patient assessment of the efficacy of DBT (OR, 4.67; range, 3.7–5.9), their feelings of hope (OR, 3.63; range, 2.9–4.4), and the increase in their feelings of confusion (OR, 0.51; range, 0.4–0.6), but it had no significant effect on physician-patient communication or the decision-making process.
Level of education proved to be an important covariate in determining the majority of the responses: it had a highly significant trend effect on the way in which patients chose a treatment, as well as on their opinions and their hopes of being cured. In particular, patients with lower levels of education were more likely to trust the advice of their physicians than they were to trust scientific progress when choosing a treatment (OR, 2.57; range, 1.7–4.0), were more convinced about the efficacy of the DBT (OR, 2.56; range, 1.4–4.6), and expressed greater hopes of being cured (OR, 2.93; range, 1.7–4.9). Furthermore, level of education had a statistically significant effect on physician-patient communication, as demonstrated by the fact that respondents with lower levels of education declared that they had greater difficulties in communicating with their physicians (OR, 0.42; range, 0.3–0.7).
The geographic location of the patients did not influence their opinions or feelings but had a certain effect on their decision-making processes and communication with their physician. Patients in northern Italy relied more on their physicians when deciding on treatment (OR, 1.36; range, 1.1–1.6), whereas those residing in central or southern Italy declared that they had greater trust in scientific advances.
With regard to gender, male patients had a better opinion of the efficacy of the DBT (OR, 1.34; range 1.1–1.7) and were less likely to follow the advice of their physicians in choosing a treatment (OR, 0.76; range, 0.6–0.9). Gender did not have any significant effect on physician-patient communication or feelings.
The current study indicates the weight of the effect that the information provided by the media has on the opinions, feelings, and attitudes of oncologic patients. The Di Bella event offered an extraordinary opportunity to study these aspects, because it was the first to involve such an intense and publicly welcomed campaign in favor of an alternative treatment (not to mention conflicts between different government institutions) that has occurred over the last 20 years in an industrialized country.3, 4
Designed and performed during the two crucial moments of the event (at the peak of the media campaign and after the negative results of the subsequent clinical trials), our study underlines the fact that the differences between the two surveys concern patient opinions and feelings but not physician-patient communication or the decision-making process. In particular, there was a considerable decrease in the positive judgments of the efficacy of the DBT (42% vs. 14%), a decrease in the hope of being cured (53% vs. 25%), and an increase in the sense of confusion (48% vs. 65%). It is worth noting that despite the scientific demonstration of the inefficacy of the DBT and the publication and release of these results by the media, only 11% of patients were convinced that it was not effective. This indicates how difficult it is to change patient opinions simply by publicizing the results of scientific research and how difficult it is for patients to give up their hopes in a new treatment. Moreover, patients believe in good news in the media (“New Hope from Miraculous Regimen ”) but are less inclined to pay attention to bad news; therefore, it is likely that large numbers of patients will continue to seek this alternative therapy.
Our results give rise to two main subjects for consideration. The first concerns the patient's informational needs and the role played by the media in supplying the public with health information, which, during the Di Bella event, was almost exclusive (only 5% of patients were informed by their physicians). This finding was confirmed by a national poll conducted in the United States in which 75% of respondents listed television and magazines or journals as their primary sources of health news.9
The vast majority of patients with cancer want a great deal of specific information concerning their illness and treatment, and this has been confirmed recently by two studies that were conducted in two European countries.10, 11 However, there is a severe lack of health information for patients attending Italian oncology departments, as demonstrated by a recent census, and this situation is aggravated by the manipulative use of health information on the part of scientifically untrained television commentators and journalists, who encouraged the expansion of the Di Bella phenomenon.12, 13 It is necessary, as suggested previously, to create a new class of journalists who are informed better about scientific matters and to whom comments concerning health news should be exclusively entrusted.14 Our data also show that the influence of the media is prevalent among individuals with lower levels of education, many of whom use the television and radio as their main source of health information (69% of individuals with only a primary school education vs. 27% of graduates).6 Therefore, persons with lower levels of education are more vulnerable to media campaigns and should be the object of special interventions designed to use the television in a positive way, as has been suggested previously.14, 15 These effects also have been highlighted in recent reports describing the greater use of screening mammography after media information campaigns.16, 17
The other relevant aspects emerging from this study concern physician-patient communication and the decision-making process. Our data show that these are not influenced by differences in information supplied by the media but depend mainly on level of education and some sociodemographic characteristics, such as geographic origin and gender (Table 3). It must be noted that the current study was designed specifically to evaluate the influence exerted by the media, and not to explore all of the factors that may affect decision-making and communication (it also is possible that other unexplored aspects and unconscious processes play a prevalent role in determining different attitudes and behaviors). The decision-making process was investigated with emphasis on the choice of unproven therapies rather than scientifically tested therapies, the role played by hope in the therapy, and the factors that influenced patients' trust in their attending physicians; the questions exploring these issues were constructed with specific consideration of the prominent media debate of the time. However, all other things being equal, people with a limited education and with a given sociocultural background (there are very large sociocultural differences between northern, central, and southern Italy) find it more difficult to communicate with their oncologist. Therefore, it seems necessary to train physicians and nurses to take these aspects into account in their patient relationships by learning how to communicate more effectively with a lay population and adapting their behaviors accordingly.18, 19
The absence of competence is one of the main obstacles to the effective participation of patients in the decision-making process and, if this is the case, then others need to take their place in making therapeutic judgments. Providing patients with the amount of information necessary for making a correct decision (while balancing the desire and moral/juridical duty to inform without generating anxiety or conflict) is part of the art of medicine: it depends on the ability of health workers to communicate and often is very difficult.20
In conclusion, our results indicate the urgency of working to improve the way in which patients are informed and educated by the media and oncologic staff. It is necessary to develop our understanding of the most effective method for informing and educating patients, bearing in mind differences in level of education and sociocultural background, because it is by affecting opinions and behaviors that information may influence health status. Finally, we also believe that it is necessary to ensure that the indisputable power of the media is directed toward improving the transfer of scientific knowledge from researchers to the public, and particularly individuals with lower levels of education.
The authors thank the following participating investigators, who supervised the study at their own centers: Giorgio Cocconi, Alessandro Barilli, and Maria Angela Bella (Parma, Italy); Giuseppe Failla, Rosa Anna Aiello, and Stefano Cordio (Catania, Italy); Carlo Gatti, Guglielmo Fila, and Adriano Paccagnella (Venice, Italy); Nina Olmeo (Sassari, Italy); Mario De Lena and Antonio Mazzei (Bari, Italy); Francesco Cognetti and Salvatore De Marco (Rome, Italy); Luigi Acito, Romeo Bastioni, and Lucio Giustizi (Fermo, Italy); Cristina Iorgu (Terni, Italy); Antonio Ardizzoia (Monza, Italy); Francesco De Falco and Giuseppa Cinquemani (Naples, Italy); Mario Battistoni, Dolores Balocchi, and Gemma Barsanti (Lucca, Italy); Roberto Labianca, Paola Poletti, and Caterina Messina (Bergamo, Italy); Orazio Caffo (Trento, Italy); and Carla Selvestrel (Verona, Italy).
The authors also thank Renata Todeschini, Francesca Diodati, and Eliana Valerani for their secretarial support and Kevin Smart (Link Srl, Milan, Italy) for his linguistic help in preparing the article.
- 7Italian National Institute of Statistics (ISTAT). Introducing Italy. Rome: ISTAT, 1996.
- 8Applied logistic regression. New York: John Wiley & Sons, 1989., .
- 9Roper Starch Worldwide Inc. Americans talk about science and medical news: the National Health Council Report. Harrison, NY: Roper Starch Worldwide, 1997.
- 12Care professionals, materials and rooms devoted to informational support to cancer patients in Italian Oncology Divisions: a national survey. Presented at the 25th Congress of the European Society of Medical Oncology, Hamburg, Germany, October 13–17, 2000. Ann Oncol. 2000; 11 Suppl 4: 148–149., , , et al.
- 15Mass media interventions: effects on health services utilisation. Cochrane Database Syst Rev. 2000; 2: CD000389., , , , .
- 20Doctor-patient communication: the lifeline to comprehensive cancer care. In: PerryMC, editor. American Society of Clinical Oncology educational book (Spring 1998). Philadelphia: W.B. Saunders, 1998: 195–201..