Table 2.

Categories for data extraction

	Category	Category options
1	Type of study	In silico, in vitro, in vivo
2	Species	For example, human, rabbit, rat, mouse
3	Direct measurement of pain	Yes or no; if so, which method was undertaken? Options: pain behaviour in animals, patient’s pain assessed in clinical practice, Visual Analogue Scale, Numerical Rating Scale or sensory testing
4	Direct interaction*	Yes or no; if not, which cells or molecules were investigated? Options: neurons, fibroblasts or cytokines. How was the direct interaction measured? Options: in the same experiment, in same paper or only mentioned in the text, but not explored experimentally
5	Experimental technique	For example, histological staining, western blot, quantitative PCR, PCR, ELISA, bulk RNA-seq, FACS, electrophysiology, animal behaviour, Ca2+ imaging
6	Mention of blinding	Yes or no
7	Mention of randomisation	Yes or no
8	Mention of power calculations	Yes or no
9	n numbers	n number used per group
10	P values	P value for individual experiment
11	Quality score†	0–3 (0, not qualified to judge; 1, low; 2, average; 3, high)
12	Disorder	For example, rheumatoid arthritis, osteoarthritis, tendinopathy
13	Type of intervention	For example, which model was used? was there an experimental manipulation, for example, antitumour necrosis factor?
14	Note	Short summary of experimental result

*Direct interaction means that both fibroblasts and neurons/pain were assessed, detected or measured in a single experimental output, for example, if the both neurons and fibroblasts were counted in the same tissue sections. Cases in which one cell type was manipulated and the other one was evaluated were also considered to be direct interactions.

†In addition to the objective quality scores recorded as items 6 to 10, we assigned a subjective quality score based on our experience of a particular experimental technique, assessing for instance, the quality of an image, extreme variability in the data pointing towards a possible lack of power, or lack of controls.