Table 2

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study weaknesses
Koes BW et al, Netherlands, 1991	5 RCTs all comparing cervical manipulation to a control	Review	Recovery	3 studies concluded manipulation better than collar and analgesics.	Studies rated poor-moderate on methodology scoring
	Total number of patients combined 111			2 studies showed no significant difference between manipulation and diazepam, a collar and transcutaneous nerve stimulation
Cassidy JD et al, Canada, 1992	100 patients with mechanical neck pain,	RCT	Mean improvement in pain as measured by visual analogue scale	17.3 (SD19.5) v 10.5 (SD14.8)	Only one treatment with pain score repeated immediately afterwards
	One manipulation v mobilization				No long term follow up
					Manipulation was not carried out by an osteopath or chiropractor
Koes BW et al, Netherlands, 1993	29 patient with neck pain	RCT	Subjective pain measurement and physical function over 12 months	Both groups improved without any statistical difference between groups	Patients were also assigned to receive treatment by their GP and placebo. These results were not reported Small numbers
	physiotherapy (n=17) v manual therapy n=12 (manipulation and mobilisation)
Skargren EI et al, Sweden, 1996	70 patients with neck pain	RCT	Subjective pain; Function, general health	Significant improvement in pain, function and general health in both groups	No breakdown of results between back and neck pain patients given for the last 2 outcomes
	physiotherapy (n=29) v chiropractic (n=41)		Sick leave	No difference
			Fulfillment of patient expectation.	41% of chiropractic group v 24% physiotherapy group
			All measured until 6 months
Verhoef MJ et al, Canada, 1997	106 patients with neck pain and 88 with neck and back pain.	Prospective cohort	Disability (Neck Disability Index)	Baseline mean score 23.5, mean score at 6 weeks 13.3 p<0.001	No control group
	All had chiropractic manipulation
Woodward MN et al, UK, 1996	28 patients with chronic “whiplash” syndrome.	Retrospective cohort	Disability at baseline and following treatment	26/28 patients had reduction in disability	Follow-up time period not specified
	All had chiropractic manipulation				Disability classified by either a chiropractor or by an orthopaedic doctor over the phone
					No control group
Jordan A et al, Denmark, 1998	119 patients with neck pain for more than 3 months	RCT	Self reported pain and disability	All treatment modality groups had improved pain levels and disability scores	All groups underwent intervention.
	intensive training physiotherapy v chiropractic treatment		Medication use	All groups progressively reduced analgesic intake	No comparison with natural progression of injury
			All of above measured until 12 months.
Giles LGF and Muller R, Australia, 1999	33 patients with neck pain for at least 13 weeks,	RCT	Disability (Neck Disability Index) Measured at 4 weeks	Manipulation group median index score reduced by 10 points p=0.001	Very small numbers particularly in acupuncture and medication groups
	chiropractor manipulation (n=18), acupuncture (n=7) or medication (n=7)			No statistically significant reduction in acupuncture or medication group	Numbers within table referring to number of patient with neck pain do not add up
			Pain as measured by visual analogue scale	Manipulation group median pain score reduced by 1.5 points p=0.002
				No statistically significant reduction in acupuncture or medication groups
McMorland G and Suter E, Canada, 1999	61 patients presenting to chiropractic with neck pain. All received chiropractic manipulation	Retrospective cohort	Neck Disability Index score	Reduction in score after 4 weeks of treatment	No statistics displayed
					244 patients who did not complete therapy because they got better or worse were excluded!
					No control group