Table 4. Adverse event rates in a range of study types and locations.

Adverse Event	Proportion	Rate	Study participant mean or median age (years)	Study type	Reference
Arthralgia
Non-severe	71/123*	58%	28	Prospective observational	This study, Indonesia, 2013
	44/165†	27%	31	Randomised trial	Burman et al, Africa and North America, 2006 [47]
Severe‡	13/519	2%	44	Retrospective observational	Schaberg et al, Germany, 1996 [5]
Nausea
Non-severe	42/129	33%	28	Prospective observational	This study
	15/165	9%	31	Randomised trial	[47]
	1/24	4%	33	Randomised trial in multi-drug resistant TB	Diacon et al, South Africa, 2009 [48]
Severe	5/519	0.9%	44	Retrospective observational	[5]
Rash
Non-severe	47/171	27%	28	Prospective observational	This study
Severe	33/519	6%	44	Retrospective observational	[5]
	18/430	4%	40	Retrospective observational	Yee et al, Canada, 2003 [4]
Vomiting
Non-severe	23/142	16%	28	Prospective observational	This study
	15/165	9%	31	Randomised trial	[47]
	2/24	8%	33	Randomised trial	[48]
Diarrhoea
Non-severe	14/153	9%	28	Prospective observational	This study
	6/165	4%	31	Randomised trial	[47]
	1/24	4%	33	Randomised trial	[48]
Temporary or permanent cessation of TB medication due to adverse event	2/200	1%	28	Prospective observational	This study
	35/332	11%	31	Randomised trial	[47]
	0/47	0%	33	Randomised trial	[48]
	121/519	23%	44	Retrospective observational	[5]
	51/1149	4%	36	Retrospective observational	Gulbay et al, Turkey, 2006 [6]

^* For the current study, denominators are the subset of the 200 patients who had ≥6 follow up visits, and did not have that symptom at baseline

^† For randomised trials, adverse event rates are given for the standard-treatment arm where a novel comparison arm was used

^‡ Severe = requiring drug cessation