The influence of study population and definition of improvement on the smallest detectable change and the minimal important change of the neck disability index

Table 1 Previous studies presenting estimates of SDC and MIC ^a

Study	NP/CR^a	Population^a	N	NDI (T₀)	GPE	Stable	SDC	MIC (sens/spec)
Cleland[9]^b	NP	acute + chronic	137	32.2-35.7	15 pt	-3 to +3	11.6*	9.5 (0.83/0.72)
Young[14]^b	CR	average 4 weeks	165	24-25	13/15 pt	-1 to +1	17.9*	8,5 (0.62/0.79)
Young[13]	NP	unclear	91	15.8-18.0	15 pt	-2 to +2	12.1*	7.5 (n.r.)
Cleland[9]^b	NP	acute + chronic	137	32.2-35.7	15 pt	-2 to +2	8.1	7.0
Cleland[8]	NP + CR	average 2 weeks	38	21.9-20.7	15 pt	-3 to +3	12.0	7.0 (0.52/0.59)
Young[14]^b	CR	average 4 weeks	165	24-25	13/15 pt	-2 to +2	15.9*	6.5 (0.57/0.67)
Jorritsma[15]	NP	chronic	76	21	7 pt	-1 to +1	8.4	3.5
Pool[10]	NP	acute + chronic	183	14.5	6 pt	2 to 4	10.5	3.5 (0.90/0.70)
Vos[12]^c	NP	acute	187	13.0-16.0	7 pt	3 to 5	7.62	n.r.^a
Trouli[11]^d	NP	acute + chronic	65	n.r.	15 pt	-3 to +3	3.03	n.r.^a

Notes:
^aNP = neck pain, CR = cervical radiculopathy, n.r. = not reported, chronic neck pain > 6 weeks;
^bCleland 2008 and Young 2010 reported two different analyses, these analyses are represented in the table as separate studies;
^cVos et. al. reported a different SDC of 1.66 in a published paper[16]. The correct SDC reported in his dissertation[12] is presented;
^dTrouli et. al. reported a SDC of 1.78 but this value does not agree with the Bland & Altman plot presented. We present a SDC estimated from the Bland & Altman plot;
*reported SDC(90) was changed to SDC(95) (SDC(95) = 1.96 × SDC(90)/ 1.65);

ISSN: 1477-7525