With special thanks to Dr. Edward Tufte and his many books on this subject.
Another article lands on my desk for editing. I hold my pen like a sword, poised to pierce every passive sentence I spot. I hate passive sentences. Passive sentences are effects without causes. In a passive sentence there is no actor. We don’t know who made the plan, who did not follow up, who performed the surgery. It’s immaculate conception! A depressingly large percentage of writers think that a passive voice conveys a tone of authority. It doesn’t. Writers who use the passive voice, whether they realize it or not, are avoiding responsibility by describing an effect without a cause or actor.
No doubt, the preceeding lecture sounds uncomfortably familiar to a fair number of former freshman English students. But it should also ring a bell among surgeons because there also exists a passive voice in clinical studies and statistical reasoning.
I suppose I attend roughly a six or seven surgeon society meetings a year. I spot the passive voice in many, maybe most podium presentations. Perhaps I’m attending the wrong meetings. At any rate, some years back, I took a famous surgeon to task on these very pages for his anti-causal, jumble of effects without causes presentation. He may not have appreciated the criticsm but perhaps his next talk was sharper.
Audiences need to spot causality-from-nowhere in podium presentations. And then stand at the microphone and ask questions about causality. “So Dr. Famous, you gave data about x, y and z but you did not describe the cause associated with this effect. What did we learn that is relevant to my practice or this particular disease state?”
I can easily recall PowerPoint presentations where the presenting physician shoveled loads of information at the audience but had no ideas. Many times we hear the passive voice (no actor, no person doing or not doing something) in verbal reasoning and then sit back as the speaker uses techniques like data mining, factor analysis or multi-dimensional scaling to reduce data to digestible bites and never mentions a cause or causal model.
The lesson for speakers and audiences is to remember that data crunching is the means to a goal, which is evidence about a causal process. When researchers at the podium do that, then their talk becomes one of those great and too rare moments of clarification, insight and, yes, beauty.
Dr. Famous and Substituting Enthusiasm for Evidence
A very important Boston surgeon visiting the University of Arizona’s medical school a few years back gave a great treatise on a large number of patients who had undergone successful heart valve surgery. At the end of the lecture, a young student at the back of the room timidly asked, “Do you have any controls?” Well, the great surgeon drew himself to his full height, hit the desk, and said, “Do you mean did I not operate on half of the patients?” The hall grew very quiet then. The voice at the back of the room very hesitantly replied, “Yes, that’s what I had in mind.” Then the visitor’s fist really came down as he thundered, “Of course not. That would have doomed half of them to their death.” The room was quiet and one could scarcely hear the small voice ask, “Which half?”
(Paraphrased from a story told by Dr. E.E. Peakcock, Jr., University of Arizona College of Medicine; as quoted in Medical World News (September 1, 1971), p. 45. )
In 1966, Drs. Grace, Muench and Chalmers (some of the most prolific proponents of evidence-based medicine) published a meta analysis of 53 published papers evaluating portcaval shunts for esophageal surgery (“The Present Status of Shunts for Portal Hypertension in Chirrhosis” Gastroenterology 1966, 684-691). The investigators rated each study according to its level of enthusiasm of the findings for the surgery and cross referenced that against the quality of the research design.
The researchers defined quality of research design as the level of random assignment of patients to treatment or control. Poor quality studies did not compare the treatment group with any proper control. The better quality research designs were those which assigned patients randomly to the treatment or the control group.
Of the 53 published studies, 6 were randon control treatment (RCT) and 47 were not. None of the RCT studies were enthusiastic about the surgical intervention. Not so for the 47 studies lacking valid controls. Thirty-four of those studies expressed marked enthusiasm for the surgery. Or put another way, 72% (34 of 47) of the non-randomized studies endorsed a surgical procedure that the gold standard studies did not.
If nothing else, this offers strong support for more meta-analysis and, again, pulling the medical reality—the cause, in other words—out of the descriptive statistics that pass for data.
One last point here: how often is the first published study testing a new technology the study that provides the strongest evidence that will ever be found for that treatment? (I’m thinking about you BAK).
That tendency (noted in the paragraph above) is why we were so impressed with the SAS meeting two weeks ago. The organizers of the program made a special effort to organize more than the usual number of Level One (highest level of RCT studies) studies and there were several long-term (up to five years) RCT follow-up studies with high numbers (500 to 800) of patient participants regarding motion preservation implants. The strong message of those eighth, ninth and even tenth study of motion preservation is that these new surgeries are evidence-based improvements over fusion surgery for many patients.
Bullet Points, PowerPoints and Points of View
Let’s talk about your treatment. And use sentences with subjects, verbs, predicates and nouns and employ the time-tested technique of combining information sequentially to form paragraphs with compete thoughts.
Or, alternatively, we can use a PowerPoint presentation to talk about your treatment.
Which approach encourages a thoughtful exchange of information and a mutual interplay between speaker and audience? Which approach amounts to a sales pitch with bullet points?
Microsoft’s PowerPoint is the fast food of information. To quote Dr. Edward Tufte, Profesor Emeritus at Yale University where he teaches courses in statistical evidence, analytical design and political economy, “PowerPoint comes with a big attitude. It forces presenters to have points, some points, any points. Audiences, therefore, endure a relentless sequentiality, one damn slide after another. Information stacked in time makes it difficult to understand context and evaluate relationships. Visual reasoning usually works more effectively when the relevant evidence is shown adjacent in space within our eyespan. This is especially the case for statistical data, where the fundamantal analytical task is to make comparisons.” (From “Beautiful Evidence” by Edward Tufte, 2006, Graphics Press LLC, pp: 158-159).
In order to, in effect, overcome the “attitude” of PowerPoint, speakers must have strong content, self-awareness and an analytical style that neutralizes or avoids the PowerPoint style. Otherwise, the danger with PowerPoint is that by reducing information to five to seven bullet points per slide, statistical reasoning becomes almost impossible and verbal and spatial thinking is perhaps fatally weakened.
To quote Dr. Tufte again, PowerPoint sets up “a dominance relationship between speaker and audience, as the speaker makes power points with hierarchical presenatation bullets to passive followers.” In a typical PowerPoint slide, the lack of space for words pushes presenters to create imprecise statements or slogans if not outright clichés.
What’s a good antidote for the ubiquitous PowerPoint presentation? A paper handout. Within the slides themselves, try to provide content-rich graphics and a clear context for data with comparisons offered. The key, I think, is to create the conditions in a meeting for several modes of learning and interplay between speaker and audience.
Fair Enough, What’s the Point?
When it comes to medical technologies, evidence is the best way we know of for answering complex questions involving biology, patient, surgeon, and product variabilities and interactions. But the manner in which presenters are delivering these 10 or 15 minute flash cards of information is not delivering, in fact, beautiful evidence. Presenting physicians and their audiences, together, need to re-commit to linking effects to causality and to raising the quality, relevance and integrity of podium presentations across the board.
In short, ruthlessly cut out those passive sentences and bullet point evasions. Because, in fact, there are some veryworthy technologies, procedures and patients who would benefit.
