Skip to main content

Table 10 Cross-tabulations of potential predictors vs. good outcome at the 4thvisit for 464 study subjects in a study of 731 chiropractic patients treated for low back pain.

From: The Nordic Subpopulation Research Programme: prediction of treatment outcome in patients with low back pain treated by chiropractors - does the psychological profile matter?

Variables Subgroups P-value
Demographic background and life-style variables   
Age *Max. 40 yrs vs. older 0.6
Sex *Men vs. women 0.5
Type of work *Sitting vs. walking/standing vs. mixed hard/light work vs. physically hard work 0.4
Smoking *No never vs. stopped/sometimes vs. daily 0.4
Smokeless tobacco *No vs. yes 1.0
Treatment reimbursed *Not or some reimbursed vs. totally reimbursed 0.001
Description of low back pain   
Leg pain *No vs. yes 0.8
Pain intensity past 24 hrs *None/weak vs. moderate vs. severe/unbearable 0.8
Duration of pain at base-line *1-7 days vs. 8-14 days vs. >14 days 0.000
Total duration of pain past year *Maximum 30 days vs. >30 days 0.000
Pain in other parts of spine past year *No vs. yes max. 30 days vs. yes >30 days 0.000
General health *Very good/good vs. OK vs. rather bad/very bad 0.03
Outcome at 4th visit *Definitely better vs. probably better/unchanged/probably worse/definitely worse/missing data 0.02
Pain intensity past 24 hrs at 4th visit *None/weak vs. moderate vs. severe/unbearable 0.06
Psychological profile   
Anxiety **No vs. borderline vs. yes 0.4
Depression **No vs. borderline vs. yes 0.3
Cognitive anxiety ***Continuous data tested as mean values with 95%CI NS
Escape avoidance ***Continuous data tested as mean values with 95% CI NS
Fearful thoughts ***Continuous data tested as median values with 95% CI NS
Physiological symptoms
And signs of pain
***Continuous data tested as median values with 95% CI NS
  1. P-values smaller than 0.05 have been written in bold and the corresponding variables have been included in further multivariate analyses
  2. * Chi-square test
  3. **Logistic regression
  4. *** Test for trend, non-significant because confidence intervals overlap