gms | German Medical Science

Medical statistics/biostatistics/biometry is a specific branch of biomathematics dedicated to the scientific analysis of medical data. It comprises all methods for the processing and analysis of medical data [1]. As “science of chance”, it is the backbone for new scientific insights and research [2] and therefore plays a crucial role in evidence-based medicine [3].

In addition to the broad topics of research and evidence-based medicine, there are further areas in the professional practice of physicians and dentists in which statistical skills are helpful. Examples for this include access to clinical policies, informing patients about risks, evaluating marketing material from pharmaceutical and medical devices companies, or interpreting results from screening tests [4].

However, the subject is usually not very popular among medical students. The term “statistics” reminds many of the school subject mathematics and thus possibly of negative school experiences. Many medical students even described it to be an unsurmountable hurdle [5]. In a survey on the importance of the subject in medical education compared with other subjects, physicians ranked biostatistics/biometry 10^th out of 15. It was thus far behind other subjects like physiology or anatomy [6].

The attitude towards a subject can support or hinder learning and thus have an impact on whether statistical methods and concepts are understood [5]. There are a number of studies on this from non-German countries [4], [5], [6], [7], [8], [9], [10], [11], [12] showing a tendency of physicians to rate biostatistics/biometry as difficult [5], [10]. On the other hand, many physicians viewed it to be important for their work [4], [6]. Also, many reported not to have acknowledged the importance of the subject during their studies but to have found the skills helpful in practice later [7]. Only few described their education in biostatistics/biometry sufficient for their needs [6].

In Germany, the attitude of physicians or dentists towards biostatistics/biometry has not been studied yet. The aim of the current study therefore was to investigate the view of practicing physicians and dentists in Germany on biostatistics/biometry in general as well as on their biostatistical education. Due to their practical experience, they can assess whether their professional work profits from biostatistical/biometric skills. Results from our study can provide guidance for future (further) education.

Our three primary questions were:

1.

What are the perception and attitude of practicing physicians and dentists in Germany towards the subject biostatistics/biometry in general as well as with regard to their work?

2.

How do physicians and dentists evaluate their own university education in biostatistics/biometry?

3.

Does the perception of the subject differ internationally?

The respondents of this study were physicians and dentists from Schleswig-Holstein who were listed in the publicly accessible online search of the Kassenärztliche Vereinigung Schleswig-Holstein (https://www.kvsh.de) and the Zahnärztekammer Schleswig-Holstein (https://www.zahnaerzte-sh.de). The prerequisite for participation was a degree in medicine or dentistry. From the listed addresses, 1,000 physicians and 350 dentists were randomly selected in two rounds. A total of 2,700 physicians and dentists from Schleswig-Holstein were contacted.

An online questionnaire was developed to measure the perception and attitude of physicians and dentists towards the subject of biostatistics/biometry. The questionnaire was developed based on three existing questionnaires [4], [6], [7]. West & Ficalora’s [6] questions were taken directly from the study. Swift et al. [4] and Miles et al. [7] used a common questionnaire that was split and analyzed separately in two studies. The full questionnaire was available from the authors upon request.

A selection of relevant questions from both questionnaires was made regarding the following topics: attitude towards the subject of biostatistics/biometry in general, usefulness of knowledge in biostatistics/biometry for practical work, assessment of the own biostatistical/biometrical training in dental or medical school. Questions on the attitude towards the subject in general were taken from West & Ficalora’s study [6]. Questions on the usefulness of knowledge in biostatistics/biometry for practical work were developed by using the study by Swift et al. [4], while questions on the assessment of the own biostatistical/biometrical training in dental or medical school were taken from the study by Miles et al. [7]. All parts of the questionnaires from the mentioned studies [4], [6], [7] were translated from English into German and adapted to the study population in order to address not only physicians but also dentists.

Overall, the questionnaire was divided into four topics. The topics A–D are described below:

• A: Attitude towards biostatistics/biometry in general
  Topic A contained questions regarding attitudes towards the subject of biostatistics/biometry in general. This section was adapted from the questionnaire by West & Ficalora [6] and asked about general attitudes towards the subject of biostatistics/biometry, the perception of the own statistical knowledge and training, attitudes towards biostatistics/biometry regarding research, and attitudes towards biostatistics/biometry regarding evidence-based medicine. For each part statements were presented and had to be rated on a 5-point-Likert scale with verbal labeling. Agreement with each statement had to be expressed on the 5-point-Likert-scale. A total of 18 statements were given (see Tab. 2, upper part).
• B: Usefulness of knowledge in biostatistics/biometry for practical work
  In topic B, physicians and dentists were asked about the relevance of biostatistics/biometry in their daily professional work. Following the study by Swift et al. [4], nine activities from the daily work of physicians and dentists which can be associated with the application of biostatistics/biometry were listed. Respondents were asked to tick the activities they perform at work. Subsequently, they were asked to indicate for which activities they perceived biostatistics/biometry to be useful (see Tab. 4). In addition, they were asked about the usability of knowledge in biostatistics/biometry for practical work (see Tab. 2, lower part).
• C: Assessment of the own biostatistical/biometrical training in human medicine or dental studies
  Topic C contained questions about biostatistical/biometrical training during and after medical or dental school. This topic was developed based on the study by Miles et al. [7]. Physicians and dentists were asked to indicate whether they ever received training in biostatistics/biometry during medical or dental school. In addition, information about the usefulness of the subject during the studies and for the career afterwards was to be provided. Respondents were also asked whether they received additional training in biostatistics/biometry besides their medical/dental studies and whether biostatistical/biometrical skills were beneficial to their career (see Tab. 6).
• D: Sociodemographic data
  The final topic asked about sociodemographic data like gender, age, field of study, year of graduation, academic degree, workplace, and research experience (see Tab. 1).

The written/online survey method was used in form of a postal letter in combination with an online questionnaire. The survey was available via the online platform https://www.limesurvey.org. Prior to the start of the survey, a pretest with ten physicians and dentists who did not subsequently participate again in the main survey was conducted. For the main survey, 2,700 physicians and dentists were informed via postal letter about the opportunity to participate in the online-survey and were also asked to partake in it. The postal letter contained instructions for the voluntary and anonymous participation as well as information on data protection and processing time. By activating the participation button in the online portal, the physicians and dentists gave their consent to participate. They were also informed about this consent in the postal letter.

The survey was available from November 24, 2020, to March 15, 2021.

Descriptive methods were used to analyze the data. All data collected were described with appropriate measures of location and dispersion. In order to investigate differences between physicians and dentists with regard to the usefulness of biostatistics/biometry for practical work as well as with regard to their own training during medical or dental school, cross tables were constructed and descriptive p-values were determined using the χ² goodness-of-fit test or Fisher’s exact test.

In subsequent exploratory analyses, associations were modeled using logistic regressions. Gender, age, amount of reported research activity (logarithmized), attitude toward mathematics and the received biostatistical/biometrical training were used as independent variables. The following statements were modeled as dependent variables:

1.

“It would benefit my career to better understand biostatistics/biometry.”

2.

“Biostatistics/biometry is a necessary skill for clinicians not involved in research.”

3.

“My training in biostatistics/biometry is adequate for my needs.”

4.

“An understanding of biostatistics/biometry is important in my work.”

5.

“The biostatistical/biometrical training from medical/dental studies is still useful for me today.”

For the first three statements, the responses “moderately agree”, “fairly agree” and “strongly agree” were scored as agreement and the other responses were scored as disagreement. Because the effect of individual parameters rather than the goodness of overall models was relevant, parameter estimators with 95% confidence intervals were used to interpret the results, and we rated associated p-values with <5% as preliminary evidence of an independent relationship. Due to the exploratory nature of the analysis, no details are provided regarding model fit or statistical requirements.

Analyses were performed with the programs Jamovi, version 1.6, and R, version 4.1.

A total of 2,700 letters were sent by post. After subtracting neutral dropouts (letter could not be delivered or person was no longer practicing medicine/dentistry), the adjusted sample was 2,605, of whom 13.67% (356/2,605) answered the survey completely or partially. Thus, the response rate was slightly lower than in comparable studies [13], [14]. 341 individuals provided information on sociodemographic data. Accordingly, 68.04% (232/341) had a degree in medicine and 30.79% (105/341) in dentistry. 1.17% (4/341) of respondents reported a degree in both subjects. In the context of the entire evaluation, the group that studied both subjects is counted as part of the physician (not dentist) group due to its small group size. Further information on sociodemographic data is provided in Table 1 [Tab. 1].

The subject biostatistics/biometry was rated as difficult by 58.76% (208/354) of the participants. 41.81% (148/354) expressed that the subject was more difficult than other subjects in medicine or dentistry. In this regard, 21.75% (77/354) of physicians and dentists said their career would benefit from a better understanding of biostatistics/biometry.

The lowest level of agreement was found for the statements on the own knowledge and training in biostatistics/biometry. 24.58% (87/354) of participants reported the training in biostatistics/biometry to be adequate for their needs. 16.48% (58/352) rated the level of training in biostatistics/biometry in medicine/dentistry as adequate, and 7.93% (28/353) reported that their previous biostatistics/biometry courses were taught effectively.

The highest level of agreement was found for the categories of research as well as evidence-based medicine. 93.48% (330/353) indicated biostatistics/biometrics was a necessary skill for a clinician involved in research. 93.79% (332/354) of physicians and dentists rated the subject as an important part of evidence-based medicine. At the same time, a large majority of 81.07% (287/354) agreed that evidence-based medicine is important for clinical practice. The data from all participants combined is shown in Table 2 [Tab. 2]. In Table 3 [Tab. 3], the responses of physicians and dentists are presented separately.

When asked about the usefulness of biostatistics/biometry, participants were asked to tick all the activities they perform in their daily professional work. Those who indicated that they perform an activity were additionally asked to assess whether they considered the subject as useful for that purpose. This resulted in a reduced sample when asked about the usefulness of the subject with regard to the various activities (see Table 4 [Tab. 4]). Biostatistics/biometry was indicated as particularly useful for the following activities: assessing marketing and advertising material from pharmaceutical and medical technology companies (88.64%: 156/176), interpreting a screening test (87.88%: 145/165), reading research publications for general professional interest (85.17%: 247/290), using research publications to explore non-standard treatment options (86.09%: 130/151), and analyzing data from own research (91.89%: 34/37). Of all participants who reported performing these activities, 80–90% expressed that biostatistics/biometry was helpful. Least importance was given to biostatistics/biometry for clinical contact with patients (30.18%: 86/285). The results can be found in Table 4 [Tab. 4].

A difference in the perception of the usefulness of biostatistics/biometry for practical work was visible in the analysis of data for own research purposes. For this activity, dentists were markedly less likely to say that biostatistics/biometry was useful. Physicians rated biostatistics/biometry as useful for all activities more often than dentists (see Table 4 [Tab. 4]).

40.00% (138/345) of physicians and dentists indicated that an understanding of biostatistics/biometry was important in their work. 65.22% (225/345) of the participants expressed that they would like to understand more about the subject, and 52.17% (180/345) answered that a better biostatistical/biometrical training in medical/dental school would have been helpful for their current work. 20.00% (69/345) rated the biostatistical/biometrical training from before as still useful today (see Table 2 [Tab. 2], lower part). Differences between physicians and dentists were visible especially for the statement “An understanding of biostatistics/biometry is important in my work” and “Biostatistics/biometry is not important in my work”. Dentists were less likely to agree with the first statement (27.62% vs. 45.76%) and agreed with the second statement more often (22.86% vs. 11.86%) (see Table 2 [Tab. 2], lower part).

Logistic regression models to explore possible influencing factors indicated no clear relationships overall (see Table 5 [Tab. 5]). With higher levels of research activity, participants tended to indicate more frequently that their careers would benefit from a better understanding of biostatistics/biometry, and that biostatistics/biometry was also relevant for non-clinicians. Older participants were also more likely to indicate that biostatistics/biometry was relevant to non-clinical practitioners, too. Participants who received biostatistical/biometrical training were also more likely to report that biostatistics/biometry was also relevant for non-clinicians and that the biostatistical/biometrical training from their studies was still useful for them today. They also indicated more often that an understanding of biostatistics/biometry was important for their own work. The latter was also stated more often by male participants.

50.14% (173/345) of the participants stated that they had received biostatistical/biometrical training during their studies. With 66.95% (158/236), this was more often the case for physicians than for dentists with 12.38% (13/105) (see Table 6 [Tab. 6]).

Physicians who graduated more recently (graduation in 2005 to 2019) were more likely to report having received a biostatistical/biometrical training in medical school. For physicians who graduated between 1975 and 1989 this was the case for 53.95% (41/76) and 92.31% (36/39) for those who graduated between 2005 and 2019. This trend was not found among dentists. The results are shown in detail in Figure 1 [Fig. 1] and Figure 2 [Fig. 2].

Overall, 43.40% (92/212) of participants reported that the subject seemed useful at study times. Dentists tended to agree more often with this statement than physicians (52.94% vs. 42.71%). 38.79% (83/214) of physicians and dentists said that the subject had been helpful for their later career. Dentists also affirmed this statement more often (52.94% vs. 37.63%). For the previous two statements, the responses of those who answered “I did not receive a training in biostatistics/biometry” were excluded. This results in a reduced sample for both statements. According to their own information, 86.96% (300/345) received no further biostatistical/biometrical training besides their medical/dental studies. 53.04% (183/345) of physicians and dentists said they would be able to do something better if they understood more about biostatistics/biometry. Physicians were slightly more likely than dentists to affirm this statement (56.78% vs. 45.71%). With regard to additional biostatistical/biometrical training besides their studies, there was no remarkable difference between physicians and dentists (see Table 6 [Tab. 6]).

For the comparison with the three reference studies [4], [6], [7] only the responses of physicians in this study were used, since physicians were the only ones interviewed in all comparative studies.

The responses of the physicians from Germany compared with those from the US study by West & Ficalora [6] were similar on average. As an exception, a strong deviation was found in the attitude toward statement 1.5 “It would benefit my career to better understand biostatistics/biometry”. With a mean of 4.4, the physicians from the study by West & Ficalora [6] clearly agreed more frequently with this statement than the German physicians with a mean of 2.6 (SD=1.18). The results are shown in Figure 3 [Fig. 3].

In comparison with the British study by Swift et al. [4], differences were visible in the evaluation of the statement “An understanding of biostatistics/biometry is important in my work”. 79.00% (103/130) of the participants from the British study agreed with this statement. Among the German physicians, only 45.76% (108/236) gave consent to this statement.

When comparing the results with the second British study [7] it becomes apparent that with 73.00% (60/82) the participants of the study by Miles et al. rated the biostatistical/biometrical training from study times as useful for their later career more often than the physicians of our study with 42.68% (67/157).

For all other comparisons, no clear differences were found (maximum difference=8%).

Our study shows that the majority of physicians and dentists in Schleswig-Holstein rate the subject biostatistics/biometry as difficult. The physicians in this study were less likely to rate the subject as important for their career or their work compared to participants in comparative studies [4], [6], [7].

Physicians and dentists indicated to recognize the value of the subject for evidence-based medicine and research. More than 90% of participants expressed that skills in biostatistics/biometrics were necessary for a clinician involved in research, nearly one-third affirmed the necessity even for clinicians not involved in research. This was stated more frequently by those who also received biostatistical/biometrical training in medical school. Biostatistics/biometry was rated as helpful for a surprisingly large number of medical activities. These included, for example, assessing marketing materials, interpreting test results, and considering new therapeutic options. However, it was found that physicians rated the subject as useful for many activities more often than dentists.

A major part of physicians and dentists expressed dissatisfaction with their biostatistical/biometrical training from their studies. Only one-fifth rated the biostatistical/biometrical training from then as still useful today. Comparable with other studies [6], [15], [16], many reported insecurities about their own biostatistical/biometrical skills (see Table 2 [Tab. 2]). Accordingly, two thirds of the participants expressed they would like to understand more about biostatistics/biometry.

A difference between physicians and dentists was found with regard to biostatistical/biometrical training in their studies. Only 12.38% of dentists stated that they received a biostatistical/biometrical training during their studies. In contrast, 66.95% of physicians received a biostatistical/biometrical training. This discrepancy between the two professional groups can be explained by the fact that the teaching of the two study programs differs clearly. In the approbation regulations for physicians (ÄapprO) the subject was taken into account ([17], § 27 para. 1). In contrast, biostatistics/biometry was not mentioned as a subject relevant to the examination in the dental approbation regulations (ZapprO) of 1955 [18]. With the new dental approbation regulations and the introduction of the cross-section “Scientific work with a focus on medical biometry, medical informatics, literature research and evaluation, and evidence-based medicine”, though, there were innovations in this regard ([19], § 72 para. 1). Thus, an important step has already been taken to counteract the strong discrepancies in biostatistical/biometrical teaching between the two study programs.

Overall, the results of this survey show that biostatistics/biometry is considered as relevant for many activities in the medical and dental practice, but many physicians and dentists do not consider themselves well prepared for it and would like to understand more about biostatistics/biometry. Therefore, it seems reasonable to develop further training courses in biostatistics/biometry. In addition to training during studies, these should focus precisely on the biostatistical/biometrical concepts relevant to the medical/dental activities mentioned in the early practical work phase. In this context, it makes sense to link the content to clinically relevant, medical topics within the framework of evidence-based medicine. This could also increase the willingness to learn. West & Ficalora gave a similar recommendation [6]. Miles et al. also described that it would be beneficial to teach biostatistics/biometry not as a singular subject but with regard to clinically practical relevance and across disciplines [7]. Researchers also called for improving biostatistical/biometrical training in residency programs [16], [20], [21]. Considering that the majority of participants in our study stated that they had not received additional biostatistical/biometrical training besides and after their medical/dental studies, there seems to be a need to catch up in this regard in Germany as well. An increased integration into residency training could be suitable for this.

For logistical reasons, only physicians and dentists from Schleswig-Holstein were contacted in our study. But due to the standardized training, we do not assume that the results would have been systematically different in other parts of Germany. In addition, most of the participants in our study had graduated some time ago. It can be assumed that the teaching received by many physicians and dentists is therefore not up to date. However, by surveying practicing physicians and dentists, we were able to gain a better overview of the extent to which biostatistics/biometry is used in the daily work life and perceived as important. Finally, the results of our study are not sufficient to provide more precise guidelines for education and training in the subject of biostatistics/biometry. But according to the results of our study an increased integration of biostatistical/biometrical content, especially in postgraduate education, seems reasonable.

A vote by the ethics commission from October 8, 2020, has been received (AZ: 20-352). No concerns were expressed.

The authors declare that they have no competing interests in regard to this article.