Modified Needleman-Wunsch algorithm for clinical pathway clustering

Abstract

Clinical pathways are used to guide clinicians to provide a standardised delivery of care. Because of their standardisation, the aim of clinical pathways is to reduce variation in both care process and patient outcomes. When learning clinical pathways from data through data mining, it is common practice to represent each patient pathway as a string corresponding to their movements through activities. Clustering techniques are popular methods for pathway mining, and therefore this paper focuses on distance metrics applied to string data for k-medoids clustering. The two main aims are to firstly, develop a technique that seamlessly integrates expert information with data and secondly, to develop a string distance metric for the purpose of process data. The overall goal was to allow for more meaningful clustering results to be found by adding context into the string similarity calculation. Eight common distance metrics and their applicability are discussed. These distance metrics prove to give an arbitrary distance, without consideration for context, and each produce different results. As a result, this paper describes the development of a new distance metric, the modified Needleman-Wunsch algorithm, that allows for expert interaction with the calculation by assigning groupings and rankings to activities, which provide context to the strings. This algorithm has been developed in partnership with UK's National Health Service (NHS) with the focus on a lung cancer pathway, however the handling of the data and algorithm allows for application to any disease type. This method is contained within Sim.Pro.Flow, a publicly available decision support tool.

Keywords: Clinical pathways; Data mining; Lung cancer.

Graphical abstract

Fig. 1

All patient pathways displayed as a heatmap.

Fig. 2

Example of the calculation for the Levenshtein distance.

Fig. 3

Example of dynamic programming using the Levenshtein distance.

Fig. 4

Example of the calculation for the Damerau–Levenshtein distance.

Fig. 5

Example of the calculation for the Jaro distance. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Example of the calculation for the Jaro–Winkler distance.

Fig. 7

Example of the calculation for the Needleman–Wunsch distance.

Fig. 8

Example of the Needleman–Wunsch algorithm.

Fig. 9

Example of bi-gram for Jaccard distance.

Fig. 10

Example of longest common subsequence.

Fig. 11

Example of longest common subsequence.

Fig. 12

Example of modified dynamic programming algorithm.

Fig. 13

Example of modified traceback.

Fig. 14

Example of feature five.

Fig. A.15

Simplified National Optimal Lung Cancer Pathway.

Fig. A.16

Modified Needleman–Wunsch Distance Matrix for Sample 2.

Fig. A.17

Comparison of the Ten Metrics Applied to Sample 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. A.18

Comparison of the Ten Metrics Applied to Sample 2. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)