| Title | On clustering patients with low back pain |
| Authors | van der Hoef, Hanneke |
| Year | 2019 |
| Volume | Archives of Data Science, Series B 1(1) / 2019 |
| Abstract | The current study contributes to the search of identifying subgroups of patients with low back pain by using clara: clustering large applications. Different from prior studies, a dimension reduction is provided by selecting key variables found in the literature. In addition, external instead of internal validation criteria are followed. Five groups are identified, which are characterized as: (1) pain has spread down into the legs (2) acute, intense low back pain which is likely to be aggravated by work (3) acute intense low back pain, not aggravated by work, and sleeping problems (4) no (activity) limitations, good recovery rate (5) chronic (i.e. more than 3 months) low back pain with a bad prognosis. Limitations and recommendations are discussed. |