Social Media Use in COPD Patients in Germany and Switzerland

Abstract

Use of social media and the Internet has changed the information-seeking behaviour and exchange of experience and information by patients. Passive observation of such online interaction between patients (social media listening) is conducted in order to understand the burden of the disease, symptom perception, and expectations from a patient perspective. For most conditions, it remains to be established how representative the social media user community is for the overall patient population. In this study, we describe internet and social media use in a population of 570 COPD patients from Germany and Switzerland. This study population is a good representation of the overall patient population in Germany and Switzerland with regards to socioeconomic data. Patients were analyzed in an exploratory fashion whether usage of the Internet to obtain disease-specific information and exchanging on COPD via social media is associated with or is independent from certain socioeconomic criteria. About three-fourths of patients indicated using the Internet to search information about COPD and about a third of patients indicated using social media to exchange with others about their disease. Results indicated that among the patients using the Internet to seek information and among those sharing information via social media, patients with very severe COPD (GOLD stage 4) were overrepresented versus milder forms of the disease. Similarly, patients with more advanced educational background were also overrepresented in the groups using social media and Internet in relation to COPD. Differences in mean age were statistically significant, but surprisingly small between social media users and non-users. No relationship with regards to social media and Internet use for COPD were observed for domiciling situation and sex.

Zusammenfassung

Die Nutzung der sozialen Medien und des Internets hat die Informationsbeschaffung ebenso wie die Art und Weise, in der Patienten ihre Erfahrungen und Informationen austauschen, deutlich verändert. Die passive Beobachtung solcher Online-Interaktionen zwischen Patienten, das sog. „Social Media Listening“, soll Erkenntnisse über die Belastung durch die Erkrankung im Allgemeinen, die Wahrnehmung der Symptome und die Erwartungen aus Sicht des Patienten gewinnen. Für viele Erkrankungen ist jedoch nicht bekannt, in wieweit die Nutzergruppe sozialer Medien für die gesamte Patientenpopulation repräsentativ ist. Für die vorliegende Studie wurden 570 COPD-Patienten aus Deutschland und der Schweiz befragt. Die Studienpopulation war in Bezug auf die sozioökonomischen Daten repräsentativ für die Gesamtheit der COPD-Patienten in beiden Ländern. Das Nutzungsverhalten der Patienten in Bezug auf die Gewinnung von krankheitsbezogenen Informationen aus dem Internet und den Austausch bez. COPD über die sozialen Medien wurde exploratorisch daraufhin untersucht, ob Zusammenhänge zu sozioökonomischen Kriterien bestehen. Etwa drei Viertel der Patienten gaben an, Informationen über COPD aus dem Internet zu suchen, während etwa ein Drittel angab, sich über die sozialen Medien mit anderen über ihre Krankheit auszutauschen.

Unter den Patienten, die das Internet zur Informationsgewinnung über COPD nutzen und unter denen, die soziale Medien zum Austausch über COPD nutzen, sind Patienten mit sehr schwerer Beeinträchtigung des Atemflusses (GOLD Stadium 4) überrepräsentiert verglichen mit milderen Stadien. Ebenso waren Patienten mit höherem Bildungsabschluss in beiden Nutzergruppen überrepräsentiert. Auch wurden statistisch signifikante Unterschiede im Durchschnittsalter zwischen den Nutzern und Nicht-Nutzern von Internet und sozialen Medien beobachtet, doch waren diese v. a. zwischen den Nutzern und Nicht-Nutzern sozialer Medien überraschend klein. Demgegenüber wurde kein Zusammenhang zwischen dem Nutzungsverhalten und dem Geschlecht oder dem Wohnort beobachtet. Die Studie überschätzt möglicherweise aufgrund der teilweisen Rekrutierung über ein Online-Angebot den Anteil der Internet-Nutzer etwas, was aber auf die o. g. Ergebnisse der Überrepräsentation von GOLD 4 und höherem Bildungsniveau keinen Einfluss hat. Die vorliegende Untersuchung sollte den Auftakt für weitere Studien bilden. Diese könnten sowohl den Einfluss der COVID-19-Pandemie auf das Informationsverhalten von COPD-Patienten untersuchen als auch durch qualitative Analysen der Bedürfnisse von COPD-Patienten in Bezug auf digitale Angebote zur Entwicklung von subgruppengerechten Informations- und Selbstmanagement-Hilfen beitragen.

Introduction

COPD

Chronic Obstructive Pulmonary Disease (COPD) is a chronic and progressive disease of the airways and lungs, characterized by chronic bronchitis and/or lung emphysema. These lead to persistent respiratory symptoms, such as cough, hypersecretion of sputum, dyspnoea, and wheezing, as well as a progressive limitation of the airflow [1]. Several hundred million people are estimated to have COPD worldwide and the disease is associated with significant morbidity and mortality. It is reported to be the third leading cause of death worldwide [2]. Most estimates from large-scale studies on the prevalence of COPD in Europe are from 5–10 % [3]. The Burden of Obstructive Lung Disease (BOLD) study, based on a standardized approach and spirometry criteria for the diagnosis, estimated the prevalence of COPD in Europe at 10 %, with considerable variation between European countries [4]. The countries in which this investigation was conducted have in the case of Switzerland a lower incidence than most European countries [5] [6], and in the case of Germany at about European average [4] [7] [8].

There are both environmental and genetic risk factors associated with COPD, with cigarette smoke exposure representing the greatest risk. Among the factors that influence disease development and progression are also age, sex, and socioeconomic status [1].

Social Media

In the past decades, social media has evolved from a niche feature for small groups of the population to a mass phenomenon. According to [9], in July 2020 among the 7.79 billion people in the world, 4.57 billion (59 %) were Internet users, and 3.96 billion (51 %) were active social media users. This widespread use of the Internet and social media has also changed information-seeking behaviour of patients [10] [11], and has led to an increasingly active use of the online environment to interact with other patients suffering from the same or similar conditions and the general public [12]. Patients use social media platforms such as online forums, Twitter, blogs, etc. to participate in discussions and to express themselves freely, in an open and uninhibited way [12]. However, a high degree of variability in Internet and social media use has been reported in patients affected by different diseases. A review indicated the percentage of patients using social media in conjunction with their condition was ranging from 51 % in patients with orthopedic disorders to 99.3 % in gynaecological conditions [13]. To our knowledge, no data on Internet and social media use have been reported for COPD patients to date. This work was conducted as an explorative analysis of Internet and social media use in COPD in the two aforementioned central European countries, Germany and Switzerland.

Social Media Listening & Research Question

Passive observation of online interactions of people on social media is referred to as social media listening (SML). A recent study from Cook and co-workers [14] analyzed a total of 849 posts originating from patients with COPD or their caregivers from Australia, Canada, South Africa, the United Kingdom and the USA, in order to understand the burden of the disease, symptom perception, and expectations from a patient perspective. Studies like this can only report what users have written in online posts, which leaves the question open, to what extent findings from SML would be representative for the overall COPD patient population, or whether certain parts of the patient community were over- or underrepresented in the group of patients actively sharing information/seeking exchange on social media. Therefore, COPD patients in Germany and Switzerland were invited to participate in a survey, assessing both their use of the Internet for obtaining information about COPD, as well as use of social media to exchange about their disease. Furthermore, demographic information and the GOLD stage of their airflow limitation were collected. Data were analyzed in an exploratory fashion, trying to assess whether usage of the Internet and social media in COPD patients as associated with or independent from certain socioeconomic criteria.

Materials and Methods

A seven-item questionnaire was generated, assessing the following (choices given in brackets)

1. sex (male/female/diverse[*])

2. age (in years)

3. use of the Internet to obtain information about COPD (yes/no)

4. use of social media to exchange with others about COPD (yes/no)
   if yes, name media (e. g. Facebook, Twitter, WhatsApp, Instagram, Internet forum...)[*]

5. highest level of education
   (no secondary education completed/secondary education completed without subsequent formal professional training/ secondary school with subsequent formal professional training/advanced secondary education/ tertiary education completed or advanced professional training completed)

6. housing/surrounding
   (large city > 100,000 inhabitants/small city 10,000–100,000 inhabitants/communes < 10,000 inhabitants)

7. last assessment of pulmonary function (FEV₁ percent predicted)
   (80 % or more, 79–50 %, 49–30 %, less than 30 %)[**]

The questionnaire was made available in two versions, one for Germany and one for Switzerland, which were identical, expect for the names of the education levels, which were adapted for country-specific language. Questionnaires were printed and simultaneously made available on the Internet in both country-specific versions (http://studienprojekt-copd.de). Medical practices and hospital ambulant centers in Germany and Switzerland were contacted and asked to place the questionnaires and a letter containing the Internet address and a Quick Response code (QR-code) into their waiting areas. Furthermore, patient organizations (national and local) and the Lung Information Service (a patient information platform operated by the Helmholtz Zentrum Munich) were asked to support the study by disseminating the information to their members and subscribers, respectively.

Questionnaires could be returned by mail (postage-free return envelopes), e-mail, fax, or by filling in an online form on the project’s Internet homepage. The intention of offering both paper-based and electronic pathways, was to avoid a bias for computer-literate patients.

The questionnaires were collected anonymously and contained no patient identifiers or links to the individual sites or groups they were disseminated through. They could solely be distinguished by country, as there were country-specific versions for Germany or Switzerland (differing in the names of education levels). Postal envelopes were destroyed to avoid traceability by postal stamps, sender information from fax or e-mail responses were deleted immediately upon receipt (questionnaire printed out, address information deleted, and original e-mail or fax destroyed), where applicable. For online responses, the Internet Protocol (IP) address was retained by the provider, but not transmitted to the study team, and was deleted after 24 hours. Taken together, all these measures ensured that the data generated from the questionnaires was fully anonymized and could not be traced back to an individual. Therefore, the project did not require approval by an Independent Ethics Committee (IEC) according to German and Swiss legislation. To confirm this, the project was submitted to four IECs in the regions where data was primarily generated. All four IECs confirmed that an approval was not required (Landesärztekammer Baden-Württemberg, Landesärztekammer Bayern, Landesärztekammer Hessen, Ethikkommission Nordwestschweiz).

Three hospital ambulatory services in Switzerland, four medical practices and seven local patient organization groups in Germany, and the Lung Information Service in Germany distributed the questionnaires and/or the link and QR code to the Internet address to their patients/members/subscribers.

Data were analyzed using the Datatab software (Datatab e.U., Dr. Mathias Jesussek, Graz, Austria), functionalities Hypothesentest-Rechner (Premium) and Deskriptiver Statistik-Test Rechner (Premium), as available from the company homepage as online tools on 04-September 2020. The variable age was assessed using an independent samples t-test, all other variables assessed for their relationship to the use of the Internet or social media in conjunction with COPD were conducted as Chi²-tests. Statistical significance was accepted for p-values less than 0.05 (p < 0.05).

Of note, percentages in the following analysis are based on the total number of returned questionnaires (n = 570); the sum of responses may be less than 570 or 100 %, though, where answers to individual questions remained/were left blank (un-)intentionally.

Results

Responses to the questionnaire items are summarized in [Table 1]. In total, 570 COPD patients (301 male, 268 female, 1 diverse) answered the questionnaire between August 2019 and May 2020. For further statistical analyses on sex, the one diverse patient has been excluded due to an insufficient number of observations, and tests were run based on a dichotomous male/female criterion. The mean age of responders was 66 years (median 67, with a range of 66 years; minimum 21 to maximum 87 years). Patients above 65 were the largest age group: below 40 years (n = 7, 1.2 %), 40–54 years (n = 68, 11.9 %), 55–64 years (n = 152, 26.7 %), 65–79 years (n = 298, 52.3 %), and patients age 80 and above (n = 45, 7.9 %). Most of the patients domiciled in small cities (10,000–100,000 inhabitants) (n = 198, 34.7 %), followed by communes with less than 10,000 inhabitants (n = 188, 33.0 %) and last in large cities (> 100,000 inhabitants) (n = 184, 32.3 %). Nine patients (1.6 %) did not complete any secondary education, 54 patients (9.5 %) finished secondary school without subsequent formal professional training, and 344 patients (60.4 %) completed secondary school with subsequent formal professional training, 106 patients (18.6 %) completed an advanced secondary education and 53 (9.3 %) patients finished a tertiary education or received an advanced professional training degree. The categorization of lung function impairment followed the GOLD stage criteria, based on percent predicted FEV₁ from forced spirometry tests. The largest group reported GOLD stage 2 (FEV₁ results of 79 % to 50 % of the predicted FEV₁, n = 240, 42.1 %), followed by GOLD stage 3 airflow impairment (FEV₁ 49 % to 30 % predicted), (n = 185, 32.5 %), GOLD stage 4 (FEV₁ below 30 %, n = 65, 11.4 %), and GOLD 1 (FEV₁ above 80 % predicted, n = 51, 9.0 %).

Tab. 1 Responses to questionnaire per question/category and descriptive statistics for age. All percent figures refer to total number of participants (n = 570). Abbreviation y = years.

Responses (by Question/Category)	Result
Age
Total number of responses	570 (100 %)
Mean	66 y
Median	67 y
SD	10.1
Minimum	21 y
Maximum	87 y
Quantile Q1	60 y
Quantile Q2	67 y
Quantile Q3	73 y
Sex
Total number of responses	570 (100 %)
Male	301 (52.8 %)
Female	268 (47.0 %)
Diverse	1 (0.2 %)
Domicile
Total number of responses	570 (100 %)
Large city	184 (32.3 %)
Small city	198 (34.7 %)
Small community	188 (33.0 %)
Highest education completed
Total number of responses	566 (99.3 %)
No response	4 (0.7 %)
No secondary education	9 (1.6 %)
Secondary without vocational training	54 (9.5 %)
Secondary with vocational training	344 (60.4 %)
Advanced secondary	106 (18.6 %)
Tertiary education	53 (9.3 %)
FEV1 % predicted
Total number of responses	541 (94.9 %)
No response	29 (5.1 %)
> = 80 %	51 (8.9 %)
79 % – 50 %	240 (42.1 %)
49 % – 30 %	185 (32.5 %)
< 30 %	65 (11.4 %)
Use of Internet to obtain information about COPD
Total number of responses	567 (99.5 %)
No response	3 (0.5 %)
Internet use	400 (70.2 %)
No Internet use	167 (29.3 %)
Social media use to share information about own COPD
Total number of responses	567 (99.5 %)
No response	3 (0.5 %)
Social media use	169 (29.7 %)
No social media use	398 (69.8 %)

Overall, 400 patients (70.2 %) indicated using the Internet for searching information about COPD, 167 patients (29.3 %) did not use the Internet for this purpose. The number of patients using social media to exchange with other users about their disease was 169 (29.7 %) in total, while 398 (69.8 %) patients reported non-use of social media for this purpose.

In total, 434 of 570 questionnaires (76.1 %) were returned by patients from Germany, and 136 (23.9 %) from patients in Switzerland. In Switzerland, most patients were contacted during their visit to the hospital ambulatory services and used the paper-based form, while in Germany a substantial proportion of patients was reached through the online letter of the Lung Information Service and through patient groups. This led to the different distribution of sources ([Table 2]), and may also account for the differences in gender-distributions between both countries (Germany: 216 males, 217 females, 1 diverse; Switzerland: 85 males, 51 females).

Tab. 2 Returned questionnaires by source and country.

	Germany	Switzerland	Total
E-mail	1 (0.2 %)	0 (0 %)	1 (0.2 %)
Paper	225 (39.5 %)	121 (21.2 %)	346 (60.7 %)
Online	208 (36.5 %)	15 (2.6 %)	223 (39.1 %)
Fax	0 (0 %)	0 (0 %)	0 (0 %)
Sum	434 (76.1 %)	136 (23.9 %)	570 (100 %)

Patients from Germany also reported more frequently to be using the Internet to obtain information about COPD compared to Swiss patients (77.0 % of German patients, 48.5 % of Swiss patients). Similarly, German patients also reported the use of social media to share information about the state of their condition more than twice as often compared to Swiss patients (34.6 % vs. 14.0 %). Both may be indicative of the above-mentioned selection bias resulting from distribution of information through patient groups and the Lung Information Service through online newsletters.

Patients were grouped as “Internet users” (IU, n = 400) and “Internet non-users” (INU, n = 167), depending on whether they indicated using the Internet to obtain information about COPD (irrespective of their use of the Internet for other purposes). Similarly, patients sharing/exchanging information about their condition actively in social media were classified as “social media users” (SMU, n = 169), those indicating the opposite as “social media non-users”, also irrespective of their use of social media for other purposes (SMNU, n = 398); see [Table 3].

Tab. 3 Patient numbers and p-values per category. Users of the Internet to search information about COPD (IU) and non-users (INU), and patients using social media to share information about their COPD condition (SMU) and non-users (SMNU) per category. Numbers for each question may not add up to 570 (100 %) responders, as some questions were left blank by individual participants (see also [Tab. 1]).

Question/Category	IU	INU	SMU	SMNU
Age
Number of responses	400 (70.2 %)	167 (29.3 %)	169 (29.6 %)	398 (69.8 %)
Mean	64.5	69.8	64.4	66.8
Median	66	71	65	68
SD	9.929	9.628	9.088	10.347
Minimum	21	41	39	21
Maximum	87	87	87	87
Quantile Q1	58	63.5	58	60
Quantile Q3	71	77	71	74.75
95 % Cl	–7.061, –3.483		–4.266, –0.467
p-value (t-test)	< 0.001 (-5.814)		0.008 (-2.679)
Cohen’s d	0.54 (medium)		0.25 (small)
Sex
Number of responses	400 (70.2 %)	167 (29.3 %)	169 (29.6 %)	398 (69.8 %)
Male	208 (36.5 %)	93 (16.3 %)	82 (14.4 %)	218 (38.2 %)
Female	191 (33.5 %)	74 (13.0 %)	87 (15.3 %)	179 (31.4 %)
Diverse (disregarded for Chi2-test)	1 (0.2 %)	0 (0 %)	0 (0 %)	1 (0.2 %)
p-value (Chi2-test)	0.439 (0.599)		0.163 (1.944)
Domicile
Number of responses	400 (70.2 %)	167 (29.3 %)	169 (29.6 %)	398 (69.8 %)
Large city	128 (22.5 %)	56 (9.8 %)	66 (11.6 %)	116 (20.4 %)
Small city	141 (24.7 %)	56 (9.8 %)	52 (9.1 %)	146 (25.6 %)
Small community	131 (23.0 %)	55 (9.6 %)	51 (8.9 %)	136 (23.9 %)
p-value (Chi2-test with Chi2(2))	0.911 (0.187)		0.068 (5.389)
Highest education completed
Number of responses	400 (70.2 %)	163 (28.6 %)	169 (29.6 %)	394 (69.1 %)
No secondary education	6 (1.1 %)	3 (0.5 %)	4 (0.7 %)	5 (0.9 %)
Secondary without vocational training	31 (5.4 %)	22 (3.9 %)	16 (2.8 %)	37 (6.5 %)
Secondary with vocational training	230 (40.4 %)	113 (19.8 %)	86 (15.1 %)	256 (44.9 %)
Advanced secondary	87 (15.3 %)	18 (3.2 %)	40 (7.0 %)	66 (11.6 %)
Tertiary education	46 (8.1 %)	7 (1.2 %)	23 (4.0 %)	30 (5.3 %)
p-value (Chi2-test with Chi2(4))	< 0.001 (20.311)		0.015 (12.278)
p-value Fisher’s exact test	< 0.001 (21.251)		0.013 (12.342)
FEV1 % predicted
Number of responses
≥ 80 % (GOLD 1)	36 (6.3 %)	14 (2.5 %)	16 (2.8 %)	35 (6.1 %)
79 %–50 % (GOLD 2)	176 (30.9 %)	64 (11.2 %)	59 (10.4 %)	181 (31.8 %)
49 %–30 % (GOLD 3)	118 (20.7 %)	65 (11.4 %)	51 (8.9 %)	133 (23.3 %)
< 30 % (GOLD 4)	57 (10.0 %)	8 (1.4 %)	36 (6.3 %)	28 (4.9 %)
p-value (Chi2-test with Chi2(3))	0.004 (13.263)		< 0.001 (24.828)

Age differed between IU (M = 64.5 y, SD = 9.93) and INU (M = 69.8 y, SD = 9.63) statistically significant (t(565) = 5.81, p < 0.001), using an independent samples t-test. Cohen’s d for this mean difference is d = 0.54 and can therefore be considered as a medium mean difference [17]. Education also had a significant relationship with the use of the Internet to research COPD information. Patients with tertiary education (87 %) or advanced secondary degrees (83 %) were IU more frequently than those with less advanced degrees (58–67 %). Additionally they were also observed more often than expected (based on the Chi²-test assumption of unrelatedness between IU and education (Chi²-test (4) with Fishers exact test = 21.251, p < 0.001). The severity of airflow limitation also yielded a statistically significant relationship, with the most severely impacted patients (GOLD 4, FEV₁ < 30 %) using the Internet the most (88 %) and also more frequently than expected (based on the Chi²-test assumption of unrelatedness between IU and GOLD stage), followed by GOLD 2 (73 %) and GOLD 1 (72 %), while GOLD 3 patients were IU less frequently (64 %) than expected (Chi²-test with Chi²(3) = 13.26, p = 0.004). Neither sex (Chi²-test = 0.59, p = 0.439) nor the domiciling situation (Chi²-test with Chi²(2) = 0.19, p = 0.911) seemed to play a role in distinguishing IU from INU. The proportion of IUs among males was 69 %, and 72 % among females. The proportion of IUs was 71 % overall, with patients from larger cities (70 %), smaller cities (72 %) and small communities (71 %) displaying a very similar behaviour.

Age was also identified as a statistically significant factor relating to SMU (M = 64.4 y, SD = 9.09) and SMNU (M = 66.8 y, SD = 10.35), using an independent samples t-test (t(565) = 2.679, p = 0.008). Cohen’s d for this mean difference is d = 0.25 and was therefore classified as a small mean difference [17]. Education also had a statistically significant relationship with social media use (Chi²-test (4) with Fishers exact test = 12.342, p = 0.013). SMU was reported overall by 30 % of the responders, which was also the frequency reported by patients with a secondary education but no professional training (30 %). The lowest frequency of SMU was found in patients with a secondary education and professional training (25 %), which was also by far the largest group in this assessment (n = 342). The groups using social media most to share information about their COPD most were patients without any secondary education (44 %), patients with a tertiary education (43 %), and patients with advanced secondary education (37 %). There were only few patients (n = 9) in the group without any completed secondary education and frequencies reported may not be as reliable. Apart from this group, a tendency to more frequent than expected SMU in patients with higher education becomes apparent, which parallels the observations with Internet use. The severity of airflow limitation showed a strong relationship to being a SMU (Chi²-test with Chi²(3) = 24.83, p < 0.001), with the most severely impacted patients (GOLD 4, FEV₁ < 30 %) using social media for this purpose the most and also more often than expected under the assumption of independence between FEV₁ and SMU (56 %). Other GOLD stages feature less SMU at a roughly comparable level (GOLD 3–28 %, GOLD 2–25 %, GOLD 1–31 %), with only GOLD 2 showing a less than expected SMU frequency within the Chi²-test. Sex (Chi²-test = 1.94, p = 0.163) did not play a role for SMU, with the proportion of SMU among males (27 %) and females (33 %) being comparable. The domiciling situation missed statistical significance only by a small margin (Chi²-test with Chi²(2) = 5.389, p = 0.069). While an average 30 % of all patients reported social media use, the proportion was higher in patients from big cities (36 %), compared to those from smaller cities (27 %) and small communities (26 %).

Discussion

This study showed that in a population of COPD patients from Germany and Switzerland, there are socioeconomic factors, which are associated with the use and sharing of disease-related information on the Internet. To our knowledge, this is the first work assessing the use of social media for sharing disease information and of the Internet to research disease information in COPD patients. The widespread use of the Internet for obtaining disease-related information (approximately 70 % of all responders) confirms that apart from the direct interaction of physicians and medical personnel, the Internet is an important means of disseminating medical information to the COPD population. The sample assessed in this study represents – despite its limited overall size – the overall COPD patient populations in Germany and Switzerland, with regards to age, sex, educational background, severity of airflow limitation, and domiciling situation.

According to the European Lung white book [3], which is referencing data published in 2011 and 2012, the prevalence of COPD in men is higher than in women, while recent data from Germany and Switzerland indicated that both genders were similarly affected [6] [8]. All studies for prevalence of COPD in European populations assessed for the white book show a clear increase of prevalence with age. The highest prevalence occurs in patients age 65 and older [15], which is also the case in our study population. The low prevalence of GOLD stage 1 in this sample also reflects the known under-diagnosis and under-treatment of mild forms of COPD [16], for which reason this sub-population would not have been accessed by our approach, but else corresponds to the different prevalence of the GOLD stages [15].The distribution of gender, age and GOLD stages of airway impairment corresponds to the expectations for a COPD population in Germany and Switzerland [1] [3]. The patient population is distributed evenly between patients in larger cities, smaller cities, and in smaller communities, which is thus representative for the population in Germany and Switzerland. Data from the Swiss Health Observatory [8] indicate comparable prevalence of COPD in patients domiciled in cities versus those in smaller communities. Previous studies in both Germany and Switzerland have also found the presence of COPD more often reported by people with a low educational level than by those with a medium or high one [6] [8]. This distribution is also reflected in our study population. In summary, the population analyzed in our study is a good representation of the overall COPD population in Germany and Switzerland. Not unexpected, there are statistically significant differences in Internet use by age, in-line with data published by the United Nations Economic Commission for Europe for both Germany and Switzerland (UNECE) [18], indicating decreasing Internet use with higher age. However, the difference in mean age between the IU and INU groups in our sample (approximately 5 years) or SMU and SMNU (approximately 2 years) were considerably smaller than we had expected from the UNECE data. Patients with more advanced education seem to be more active in obtaining information about COPD from online sources. Whether this correlates with lesser use of other sources of information, or reflects a generally higher demand for information remains to be discovered in future research. The patients with very severe airway impairment (GOLD 4) were the ones with the highest proportion of Internet-users. We hypothesize that this observation might be explained either by the greater burden of the disease, which could lead to increased activity in learning about COPD, or by the inability to access other sources of information (e. g. patient symposia, patient groups). Patients with mild and moderate airway impairment have a comparable level of Internet-use (72 % and 73 %, respectively). The slightly lower rate of Internet users among the GOLD 3 patients (severe airway impairment, 64 %) is difficult to interpret. If this observation prevails in larger future studies, the time since COPD diagnosis or a preference for use of other means of information such as patient groups may be aspects that could be investigated. Equal use of the Internet for information purposes by male and female users is also in-line with the data reported by UNECE for Germany and Switzerland [18]. Despite differences in access to broadband Internet between larger cities and small communities in the countryside [19] [20], for example, the domiciling situation did not appear to impact the choice to access information about COPD via the Internet. It may be due to the nature of the information (predominantly texts and pictures or very short video clips), which is little dependent on the quality of the infrastructure / transfer rates, that no difference was observed. Furthermore, the advantage of being able to access disease information at any time, and the possibility to search for information on topics considered embarrassing (e. g. sputum, urinary incontinence, etc.) anonymously may make Internet searches for COPD-related information equally attractive in cities and smaller community settings.

In contrast to Internet use, only a minority of some 30 % of responders indicated that they actively share information about their own COPD via social media. Similar to Internet use, a statistically significant correlation between social media use for COPD and age, education, and severity of airway impairment (GOLD-category) was found in this study. However, the means difference for age was very small (2 years), so that in essence there is no clinical bias for younger responders to be expected when listening to social media. Notably, the distribution between SMU and SMNU with regard to severity of airflow limitation (GOLD classes) seems to mirror that of Internet users: in GOLD stages 1–3, the fraction of SMU hovers between 25–31 %, but in GOLD 4 increases to 56 % of responders in that category. Similar to Internet use, one may hypothesize that the greater burden of the disease and reduced access to other forms of exchange such as patient groups or patient symposia may trigger this effect. Of note, while statistical significance for a correlation of the domiciling situation to social media use was missed by a smaller margin, it was notable that social media use to exchange about COPD was numerically more pronounced in large cities, compared to both smaller cities and small communities.

A weakness of the study is the recruitment of patients through the online platform of the Lung Information Service. While originally collection of questionnaires at the Lung Information Service’s “patient day” was planned in March 2020, the event was cancelled due to the first wave of the COVID-19 pandemic in Germany. In lieu, the link to the online questionnaire was mailed out with the service’s newsletter, thereby biasing the population in Germany partially for a more internet-affine subpopulation. In contrast, the vast majority of patients in Switzerland responded on paper during visits to the practice or clinic, respectively. We interpret the differences in internet use (77.0 % of German patients versus 48.5 % of Swiss patients) and social media use (34.6 % versus 14.0 %) as reflecting this selection bias, rather than an actual difference in behaviour between both populations. For future research, it may be of particular interest to assess differences in an overall population from practices and clinic wards versus those groups that can be reached by online communication.

In summary, our results indicate that patients from Germany and Switzerland reporting about their COPD and related experiences in social media are a good representation of the overall patient population in these countries, despite the above described bias. Researchers should keep in mind, though, that their results may be biased for responses from patients with very severe COPD rather than milder forms of the disease, and for patients with more advanced educational background, who both may be overrepresented. While there is also a bias for age, it is considerably smaller than expected, and may not have any practical implications. [Fig. 1] is a graphic illustration of how the population of COPD patients in Germany and Switzerland represented in social media listening exercises deviates from an “ideal COPD patient” population according to our data collection. More research will be required to broaden these insights, in particular to cover further countries and regions. Moreover, our study describes data collected prior to the COVID-19 pandemic and in its early stages (first wave in Germany and Switzerland). Further investigations may also show an increased trend to using digital solutions such as the internet and social media in correlation with COPD-information gathering, sharing and personal disease management following the extended unavailability for face-to-face options during the lockdowns.

Fig. 1 Spider net diagram illustrating the deviation of the COPD patient population represented by social media users in Germany and Switzerland versus an “ideal COPD patient population” with no bias. The population defined by the exchange of experiences and information about their COPD in social media over-represents patients with very severe airway impairment (GOLD 4) and higher educational background, and to a small degree younger COPD patients. No bias was observed with regards to the domiciling situation and sex.

Conclusions and implications for future research: Beyond the immediate question to assess potential biases in information gained from social media listening in COPD patients, this study could also trigger broader and deeper research to analyze the needs of COPD patients for digital information about their disease. An understanding of the qualitative demand for digital sources for information and exchange, as well as of the penetration of digital competence in different subgroups of the large COPD patient community could support the development of targeted digital solutions for disease self-management.

Conflict of interest

FSG and JHE are full-time employees of Novartis Pharma AG, but have contributed to this work outside of their employment. This work was not funded or otherwise supported by Novartis Pharma AG. JHE is a visiting lecturer for medical device technology at Pforzheim University, Germany.

Acknowledgements

The authors would like to thank the following participating physicians and patient groups for their support: Atemwegserkrankungen Bergstraße (Mrs. Waltraud Zöller); Atmen ist Leben Selbsthilfegruppe (Mr. Dieter Schmiedel); Deutsche Sauerstoff- und Beatmungsliga LOT e.  V. – Gruppe Mühldorf am Inn (Mrs. Annette Hendl); Deutsche Sauerstoff- und Beatmungsliga LOT e. v. – Gruppe Ulm (Mrs. Silvia Thanner); Institut für Epidemiologie, Biostatistik und Prävention der Universität Zürich und mediX Gruppenpraxis Zürich (Prof. Dr. med. Claudia Steurer-Stey); Lungeninformationsdienst, Helmholtz Zentrum München (Dr. Denise Mackrodt); Praxis Dipl.-Med. Hagen Holz, Pocking; Praxis Dr. med. Christine Schneider, Schmidham, Ruhstorf an der Rott; Praxis Dr. med. Heidrun Kissinger, Karlsruhe; Praxis Dr. Heiner Steffen, Landsberg am Lech (Mr. Ronny Bauer); Selbsthilfegruppe COPD & Lunge (Mrs. Mary-Lou Schönwälder); Selbsthilfegruppe Lungenemphysem – COPD Region Nordbaden (Mrs. Birthe Gartner); Selbsthilfegruppe Offenburg (Mr. Lothar Teichert); Spital Uster, Pneumologie, Uster (ZH) (Dr. med. Alice Zürcher); Universitätsspital Basel, Pneumologie (Prof. Dr. med. Daiana Stolz; Mrs. Regula Vinzens).

The authors would like to thank Dr. Björn Walther (Jena, Germany) for support with the statistical analysis and critical review of the manuscript.

^* Following feedback from users, the questionnaire was updated after approximately 250 responses had been obtained, with the option “diverse” (referring to the third gender) for item 1, and with an explanatory statement defining social media for item 4.

^** Participants were referred to their physician in case of doubt.

• References

• 1 Global Initiative for Chronic Obstructive Lung Disease, Inc. (GOLD). Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructuve Pulmonary Disease (2020 Report). Available at (last accessed 17.08.2020): https://goldcopd.org/wp-content/uploads/2019/12/GOLD-2020-FINAL-ver1.2-03Dec19_WMV.pdf
• 2 World Health Organization (WHO). Global Health Estimates 2016: Deaths by Cause, Age, Sex, by Country and by Region, 2000–2016. Geneva, Switzerland: WHO; 2018
• 3 European Lung Foundation and European Respiratory Society. European Lung white book (Chapter 13 – Chronic obstructive pulmonary disease). Available at (last accessed 06.12.2020): https://www.erswhitebook.org/files/public/Chapters/13_COPD.pdf
• 4 Buist AS, McBurnie MA, Vollmer WM. et al. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007; 370: 741-750
• 5 Bridevaux PO, Probst-Hensch NM, Schindler C. et al. Prevalence of airflow obstruction in smokers and never-smokers in Switzerland. Eur Respir J 2010; 36: 1259-1269
• 6 Schweizerisches Gesundheitsobservatorium (Swiss Health Observatory). Chronische Bronchitis/Emphysem. Available at (last accessed 05.12.2020): https://www.obsan.admin.ch/de/indikatoren/chronische-bronchitisemphysem
• 7 Geldmacher H, Biller H, Herbst A. et al. Die Prävalenz der chronisch obstruktiven Lungenerkrankung (COPD) in Deutschland [The prevalence of chronic obstructive pulmonary disease (COPD) in Germany. Results of the BOLD study]. Dtsch Med Wochenschr 2008; 133: 2609-2614
• 8 Steppuhn H, Kuhnert R, Scheidt-Nave C. 12-month prevalence of known chronic obstructive pulmonary disease (COPD) in Germany. J Health Monitor 2017; 2: 43-50
• 9 wearesocial and Hootsuite. Digital 2020: July Global Statshot Report. Available at (last accessed 17.08.2020): https://www.slideshare.net/DataReportal/digital-2020-july-global-statshot-report-july-2020-v01
• 10 Cordoş A-A, Bolboacă S, Drugan C. Social media usage for patients and healthcare consumers: a literature review. Publications 2017; 5: 9
• 11 Giustini DM, Fraser M, Boulos MNK. Effective uses of social media in public health and medicine: a systematic review of systematic reviews. Online J Public Health Inform 2018; 10: e215
• 12 Timimi FK. Medicine, morality and health care social media. BMC Med 2012; 10: 83
• 13 De Martino I, D’Apolito R, McLawhorn AS. et al. Social media for patients: benefits and drawbacks. Curr Rev Musculoskelet Med 2017; 10: 141-145
• 14 Cook NS, Kostikas K, Gruenberger J-B. et al. Patients’ perspectives on COPD: findings from a social media listening study. ERJ Open Res 2019; 5: 00128-2018
• 15 Ntritsos G, Franek J, Belbasis L. et al. Gender-specific estimates of COPD prevalence: a systematic review and meta-analysis. Int J Chron Obstruct Pulmon Dis 2018; 13: 1507-1514
• 16 Rossi A, Butorac-Petanjek B, Chilosi M. et al. Chronic obstructive pulmonary disease with mild airflow limitation: current knowledge and proposal for future research – a consensus document from six scientific societies. Int J Chron Obstruct Pulmon Dis 2017; 12: 2593-2610
• 17 Cohen J. A power primer. Psychological Bulletin 1992; 112: 155-159
• 18 United Nations Economic Commission for Europe (UNECE). UNECE Statistical Database. Available at (last accessed: 23.09.2020): https://w3.unece.org/PXWeb2015/pxweb/en/STAT/STAT__30-GE__09-Science_ICT/02_en_GEICT_InternetUse_r.px/
• 19 BAKOM (Bundeamt für Kommunikation). Breitbandatlas der Schweiz. 2020 Available at (last accessed 28.11.2020): https://www.bakom.admin.ch/bakom/de/home/telekommunikation/atlas.html
• 20 Aktuelle Breitbandverfügbarkeit in Deutschland (Stand Ende 2019). Berlin, Germany: BMVI (Bundesministerium für Verkehr und digitale Infrastruktur); 2020

Publikationsverlauf

Eingereicht: 18. Dezember 2020

Angenommen: 12. April 2021

Artikel online veröffentlicht:
07. Mai 2021