9 and so on; all problems ranked 0 or not ranked by an expert were given 0 score.
● In case of tie in ‘a’, a problem with less number of
0 scores was ranked higher.
● In case of tie in ‘b’, for all problems getting an average score of 0, the same were ranked according to the percentage of experts including them.
● Otherwise, in case of tie in ‘b’, a problem deemed more important was ranked higher vis-a-vis reasons given by various experts to support their rankings.
● In case of tie in ‘c’, a problem deemed more impor-tant was ranked higher vis-a-vis reasons given by various experts to support their inclusions.
In the third round, a list of 14 problems ranked as per the above criteria was shared with all the experts. The reasons for their inclusion by various experts were also shared. In this round, the experts were asked to rank their top ten problems, by giving ranks from one to ten. All the experts were again asked to give reasons to support their rankings in order to gauge any creativity in their thought process. The same was planned in the subsequent rounds, if needed, to facilitate informed consensus. Finally, at the end of the third round, a ranked list of all the problems included by different experts, while excluding those not ranked 1–10 by any of the experts, was generated as follows:
● Mean rank score (MRS) was used for ranking these problems; a problem with first rank was given a score of 10, second rank 9 and so on until the tenth rank was given a score of 1; problems not included by an expert in the top ten were given a score of 0.
● In the case of equal MRS, a problem included in the top ten by higher percentage of experts was ranked higher.
● In the case of tie in ‘b’, a problem deemed more important was ranked higher vis-a-vis reasons given by various experts to support their rankings.
The top ten problems as per the ranking of the third round of the Delphi technique were analysed using statistical tests (i.e. percentage, mean rank score, stan-dard deviation, median, inter quartile range and rank). The corresponding MRS for these problems was also considered from the earlier two rounds. The main goal of this statistical analysis was to determine whether:
● Optimal informed consensus had been reached on the most vital problems vis-a-vis strategizing improvement of quality in nursing care in these wards.
● Subsequent rounds of the Delphi technique would contribute significantly towards greater consensus