Statistical Reasoning for Surgeons

Mitchell Maltenfort lurched into academic life as a computational neurobiologist before drifting into the less recherché field of biostatistics. He knows just enough to make a complete hash out of things and is creative enough to salvage them afterwards. In his brutish culture, this tradition is known as "larnin'." For tax purposes, he is employed as a biostatistician at CHOP, where he has generated risk scores for hospitalization, analyzed diagnostic variations among clinics, compared international trends in childhood mortality, and evaluated patient-reported outcome scores. Antonia Chen is the Director of Research for Arthroplasty Services at Brigham and Women’s Hospital and an Associate Professor at Harvard Medical School. She is a past president of the Musculoskeletal Infection Society (MSIS) and is an active collaborator in research studies on topics including infection outcomes and opioid use. She often travels armed with a small and extremely cute dog named Lily who is often the highlight of research meetings where Lily is in attendance. Camilo Restrepo is the Associate Director for Research at the Rothman Institute. He is known informally at RI as "Doctor Data."

1 Introduction – Why Does a Surgeon Need Statistics?

2 Interpreting Probability: Medical School Axioms of Probability

3 Statistics, the Law of Large Numbers, and the Confidence Interval

4 The Basics of Statistical Tests

5 How Much Data Is Enough?

6 Showing the Data to Yourself First – Graphs and Tables, Part 1

7 How Normal Is a Gaussian Distribution? What to Do with Extreme Values?

8 All Probabilities Are Conditional

9 Quality versus Quantity in Data

10 Practical Examples

11 All Things Being Equal – But How? (Designing the Study)

12 Binary and Count Outcomes

13 Repeated Measurements and Accounting for Change

14 What If the Data Is Not All There?

15 Showing the Data to Others – Graphs and Tables, Part 2

16 Further Reading

GLOSSARY

REFERENCES

INDEX