Outcomes of pregnant patients with low back pain undergoing chiropractic treatment: a prospective cohort study with short term, medium term and 1 year follow-up

BACKGROUND Low back pain in pregnancy is common and research evidence on the response to chiropractic treatment is limited. The purposes of this study are 1) to report outcomes in pregnant patients receiving chiropractic treatment; 2) to compare outcomes from subgroups; 3) to assess predictors of outcome. METHODS Pregnant patients with low back or pelvic pain, no contraindications to manipulative therapy and no manual therapy in the prior 3 months were recruited.Baseline numerical rating scale (NRS) and Oswestry questionnaire data were collected. Duration of complaint, number of previous LBP episodes, LBP during a previous pregnancy, and category of pain location were recorded.The patient's global impression of change (PGIC) (primary outcome), NRS, and Oswestry data (secondary outcomes) were collected at 1 week, 1 and 3 months after the first treatment. At 6 months and 1 year the PGIC and NRS scores were collected. PGIC responses of 'better' or 'much better' were categorized as 'improved'.The proportion of patients 'improved' at each time point was calculated. Chi-squared test compared subgroups with 'improvement'. Baseline and follow-up NRS and Oswestry scores were compared using the paired t-test. The unpaired t-test compared NRS and Oswestry scores in patients with and without a history of LBP and with and without LBP during a previous pregnancy. Anova compared baseline and follow-up NRS and Oswestry scores by pain location category and category of number of previous LBP episodes. Logistic regression analysis also was also performed. RESULTS 52% of 115 recruited patients 'improved' at 1 week, 70% at 1 month, 85% at 3 months, 90% at 6 months and 88% at 1 year. There were significant reductions in NRS and Oswestry scores (p < 0.0005). Category of previous LBP episodes number at one year (p = 0.02) was related to ,improvement' when analyzed alone, but was not strongly predictive in logistic regression. Patients with more prior LBP episodes had higher 1 year NRS scores (p = 0.013). CONCLUSIONS Most pregnant patients undergoing chiropractic treatment reported clinically relevant improvement at all time points. No single variable was strongly predictive of, improvement' in the logistic regression model.