to help with support, maintenance and iterative improvements of
the system independently from the vendor. This process reflects
the current discussion critiquing PD conducted by external designers
as often failing to provide the necessary level of user input over
time [34]. The experience at SN gives a concrete example of how to
start building teams that allow in depth knowledge obtained during
the design, development and implementation of an EMR system
to be contained within a hospital. This may thus used as a
starting point in considering how to ‘‘design a [socio-technical]
system that more or less then designs itself’’ ([34], p. S103) The
long term success of the strategy in allowing SN to continue with
iterative improvements of the system independently from the vendor
is still unclear however, and is something worthy of investigation
in the future.
5.2. Limitations
As with all case studies the usual limitations must be noted with
respect to problems that may arise when trying to generalize the
results in dissimilar settings [35]. In India for example many health
workers at public hospitals are members of labor unions. In that situation
the relationship between management, EMR design teams,
and health care workers may be fundamentally different than
reported in this study. The limitations with data analysis noted in
the methods section are also worth re-stating, as the somewhat limited
scope of the study during the data collection phase, and problems
in conducting further data analysis due to partial data loss,
may have resulted in a number of interesting and important issues
related to the findings not being reported in the study.