Discussion What does all this mean Implications Discussion

Discussion What does all this mean? Implications

Discussion • State major findings • Strengths and limitations – Design, technique, results • Discuss – findings re: existing information – “important” minor findings • Implications – of findings, for future research, new questions • Summary

Discussion: Problems • • Difficult to write Tend to be verbose, too long Issues not addressed by the study Sequence: does not flow Introduce new information Harp on trivial or obvious points Ignore difficult to explain results

Discussion: Do not • • • Start with history Repeat all your results Provide new data Extrapolate results Superlatives

Discussion: Make life easy • Think and plan – Sequence • Identify subheadings (4 paragraphs) – – – Salient results: few sentences Strengths and limitations Compare and contrast: Existing literature Implications Conclusions: Take home message – For yourself

Focus the discussion • Begin with the most important point • Confine to the subject studied • Focus on key issue • Provide link sentences between paragraphs: ensure flow

Strengths • Study design – Sample size, controls, variables assessed – Length of follow up • Technique – New, established but not used in subset

Limitations • • Selection bias ? Generalize, apply Drop outs Discuss unexpected findings – Do not ignore, offer explanation – If none plausible: say so

Compare/contrast • • Must: Good grasp of information Provide explanations for differences Avoid rhetoric; be diplomatic Do not hesitate to criticize – reasons, logical argument – Discuss opposing views

Implications • • Inter-relationships Alteration in clinical practice Speculate, but intelligently Geographical Financial Consider alternative explanations Take care: – Correlation does not indicate causation

Conclusions • Answer – So what? – Who cares? • Only those supported by data • Avoid sweeping statements Take home message

Is the job done? • Ask colleague to read • Get feedback • Re-work – Not till these done

Examples • Arnold and Chiari first described… 1942 • Our results indicate an increasing prevalence of… 48% in 1992 to 69% in 2003

Examples • This is the most authentic collection of data ever. • It could be argued that our data has no major clinical implications. However, …

Examples • This study surpasses all previous work in this field… • In 1971, …showed no improvement. However, they studied…methods with a low specificity. We…high specificity, … recently available, … different results.

Summary • • Break in to sections: Structure Provide a link sentence: Sequence Short, clear, key issues Limitations So what? Who cares? Take home message