Table 2.

Prioritization of user requirements using the decision matrix (example shows 4 out of 64 therapist requirements from the category ‟monitoring”).

ID	Category 1: Monitoring	Decision criteria
	Description of requirement (number of entries)	Literaturea (+ or − or ?)	Defined by majority of usersb (+ or −)	Consensus patient therapistc (+ or −)	Complexity 0= very low 10= very high	Priorityd 1=high 4=low	Notes
1e	The system must be able to monitor the intensity of phantom limb pain, so that the therapist is able to evaluate its course over time (10/10)	+ Barbin et al [8] Rothgangel et al [9]	+	+	5	1 1 1
2	The system has to record the perceived position and range of motion of the phantom limb (1/10)	+ Schmalzl et al [33] Mercier and Sirigu [34] Moseley [35] Sumitani et al [36]	-	-	8	3 4 3	Consider for clinical trial
3e	The system must enable the therapist to control the frequency and quality of self-delivered exercises (eg, video recording, text messages) (10/10)	+ Darnall and Li [11] Beaumont et al [37] MacIver et al [38]	+	+	8	1 1 2	Camera of tablet has no wide angle—poor display window
4	The system has to record the perceived difficulty of self-delivered exercises (3/10)	+ Mercier and Sirigu [34] Beaumont et al [37] Giraux and Sirigu [39]	-	-	5	3 2 3

^a+= yes, −= no, ?=unclear.

^b+=Requirement defined by >50% of users.

^c+=consensus between at least one patient and one therapist.

^d1=must have, 2=should have, 3=could have, 4=won’t have this time.

^eBased on the decision criteria and priority rating only requirements with ID 1 and 3 were defined as critical for the first prototype.