Skip to main content

Table 3  Univariate analysis of the categorized diagnosis, medication, and lab data of the non-CC and CC clusters

From: Application of unsupervised deep learning algorithms for identification of specific clusters of chronic cough patients from EMR data

  Non-CC (N = 8588) CC-1 (N = 4658) CC-2 (N = 1737) CC-3 (N = 154) p value
Diagnosis
Respiratory 2003 (23.32%) 4300 (92.31%) 1593 (91.71%) 140 (90.91%) < .0001
Endocrine metabolic 1268 (14.76%) 3514 (75.44%) 1026 (59.07%) 88 (57.14%) < .0001
Circulatory system 1687 (19.64%) 3735 (80.18%) 1110 (63.9%) 91 (59.09%) < .0001
Mental disorder 1020 (11.88%) 3133 (67.26%) 857 (49.34%) 56 (36.36%) < .0001
Neurological 805 (9.37%) 2381 (51.12%) 579 (33.33%) 38 (24.68%) < .0001
Digestive 1419 (16.52%) 2889 (62.02%) 767 (44.16%) 60 (38.96%) < .0001
Symptoms 2344 (27.29%) 3335 (71.6%) 943 (54.29%) 71 (46.1%) < .0001
Hematopoietic 336 (3.91%) 2353 (50.52%) 484 (27.86%) 20 (12.99%) < .0001
Medication
Antiasthmatic bronchodilator 1099 (12.8%) 2777 (59.62%) 898 (51.7%) 79 (51.3%) < .0001
Minerals electrolytes 1164 (13.55%) 2791 (59.92%) 704 (40.53%) 52 (33.77%) < .0001
Corticosteroids 1035 (12.05%) 2529 (54.29%) 728 (41.91%) 61 (39.61%) < .0001
Ulcer drugs 1714 (19.96%) 2789 (59.88%) 769 (44.27%) 55 (35.71%) < .0001
Lab
Blood count 3437 (40.02%) 4078 (87.55%) 1287 (74.09%) 105 (68.18%) < .0001
  1. The p values for the comparison are bracketed in the last column. Only < .0001 was listed in the last column if all p values were < .0001