Re: Data, data everywhere
The “Ballad of John Henry” tells of the legendary black American steel pin driver, John Henry, who swung a huge nine pound hammer driving railroad spikes on the Chesapeake and Ohio railway in the 1870’s. John Henry was renowned for his strength amongst his fellow workers and could drive in a steel pin into a track with a single blow instead of the usual three. The ballad centers on a competition between John and a mechanical steam powered drill, a controversial innovation that threatened to replace thousands of his fellow laborers in a quest for greater efficiency. When John Henry heard that the output of his fellow laborers were being compared with that of a steam drill, he challenged the railroad company to a contest pitting his own skill and strength against that of the stream drill to see who could lay the track the quickest.
The steam drill was positioned on one side of the track and John Henry on the other. When the signal to start was given two thousand people came to watch the event. John Henry made more progress in a shorter time than the steam drill and when he reached the finish line the steam drill was no where in sight. He won the contests but worked so hard to outperform the machine he collapsed and died in the process. The John Henry is just one example of reactivity, the phenomenon whereby individuals alter their performance or behavior to the awareness that they are being observed. The term John Henry effect, also known as compensatory rivalry, was first suggested by Robert Heinich in 1970 then further developed by Gary Saretsky in 1972 to describe the behavior. 
If we were to cast this tale into the future then computers powered by big data would be represented by the steam drill and the doctor by John Henry. Perceiving the consequences of such an innovation as threatening to their jobs, status, or traditional patterns of working, doctors may go to extraordinary lengths to outperform the opposition. However, the increased work required for victory may prove unsustainable for some and come at considerable personal cost.
Such effects may confound evaluation outcomes unless controlled for by robust experimental design e.g. adequate blinding. When such evaluations ultimately take place (and they will) we should not forget that doctors do far more than just crunch data (or lay track). Indeed, the interpersonal warmth, trust and informality which characterise most clinical consultations is where we witness the real victory of human dignity, intuition, and lateral thinking over the politically driven degradations of the machine age.
1. Belafonte, H. John Henry. 1959 https://www.youtube.com/watch?v=g6vcvYJCkic [accessed 5.2.13]
2. Saretsky G. The OEO PC experiment and the John Henry Effect. Phi Delta Kappa 1972; 53:579–581
3. Metcalfe D. Competence, administration and learning. London: MSD Foundation 1989.
Competing interests: No competing interests