Table 3.

Summary of results: effectiveness of the interventions with respect to thyroid function tests

Outcome	Study	Intervention		Direction	Large effect	Statistically significant change	RCT design	Notes on outcome measures
Single-mechanism interventions
Test numbers or rates	Adlan et al22	Guidelines		+	+	+	−	Per cent admissions offered TFTs
	Berwick and Coltin17	Education (test specific)		+	−	NR	−	Per 100 encounters
	Berwick and Coltin17	Feedback on cost		+	−	NR	−	Per 100 encounters
	Berwick and Coltin17	Feedback on yield		−	+	NR	−	Per 100 encounters
	Chu et al24	Guidelines		+	+	+	−	Per 100 ED visits
	Cipullo and Mostoufizadeh19	Guidelines		+	−	NR	−	Per discharge
	Gama et al18	Feedback		+	+	+	−	Per outpatient visit
	Emerson and Emerson27	Request form redesign		+	−	+	−
	Feldkamp and Carey28	Guidelines		+	−	NR	−	Per 1000 patients
	Grivell et al29	Feedback		−	+	NR	−
	Horn et al45	Display of cost of tests being ordered		+	−	−	−	Per 1000 visits
	Schectman et al33	Educational memorandum		+	−	+	−	Per patient
	Thomas et al10	Feedback		+	−	+	+	Per 10 000 registered patients
	Thomas et al10	Reminders		+	−	+	+	Per 10 000 registered patients
Appropriateness (compliance)	Daucourt et al25	Pocket memory card		+	−	−	+	Proportion of TFTs ordered in accordance with the guidelines
	Daucourt et al25	Request form redesign		+	+	+	+	Proportion of TFTs ordered in accordance with the guidelines
	Schectman et al33	Educational memorandum		+	+	+	−	Compliance with TSH-only strategy
	Schectman et al33	Reminders		+	−	+	−	Compliance with TSH-only strategy
Expenditure	Tierney et al35	Display of computer- generated probability estimates		+	−	−	+	Per visit
CV	Berwick and Coltin17	Education (test specific)		−	−	NR	−
	Berwick and Coltin17	Feedback on cost		+	+	NR	−
	Berwick and Coltin17	Feedback on yield		+	+	NR	−
Pattern	Emerson and Emerson27	Request form redesign		+	+	NR	−	Sought to shift to TSH and thyroid cascade
	Feldkamp and Carey	Guidelines		+	+	NR	−	Sought to shift to TSH and TSH-based algorithm
	Larsson et al28 and Mindemark and Larsson31	Years 1–2	Education	+	−	+	−	Sought to shift to TSH; and reduce ordering of TT3 and FT4 relative to TSH. Summary based on TSH/all TFTs ratios
	Larsson et al28 and Mindemark and Larsson31	Years 2–6	Education	−	−	+	−	Sought to shift to TSH; and reduce ordering of TT3 and FT4 relative to TSH. Summary based on TSH/all TFTs ratios
Multiple-mechanism interventions
Test numbers or rates	Baker et al	Education and guidelines		+	−	−	+	Per 1000 registered patients
	Daucourt et al	Pocket memory card and request form redesign		+	−	+	+	Proportion of TFTs ordered in accordance with the guidelines
	Dowling et al26	Education and feedback		+	−	+	−	Per patient visit
	Hardwick et al44	Funding policy and guidelines		+	−	NR	−
	Rhyne and Gehlbach43	Education and guidelines		+	+	+	−	Per 100 encounters
	Schectman et al	Feedback and reminders		+	−	+	−	Per patient
	Thomas et al10	Feedback and reminders		+	−	+	+	Per 10 000 registered patients
	Tomlin et al36	Education and feedback and guidelines		+	+	+	−	Per year per GP
	Toubert et al37	Guidelines and reminders		+	+	NR	−
	van Walraven et al39	Guidelines and funding policy		+	+	+	−	Summary based on decrease in the proportion of TT4 and T3RU
	van Walraven et al39	Guidelines and request form redesign		+	−	+	−	Summary based on decrease in TSH utilisation
	Vidal-Trécan et al40	Education and guidelines and request form redesign		+	−	NR	−	Summary based on the total number of TFTs
	Willis and Datta41	Education and guidelines		+	+	+	−	Per admission
	Wong et al42	Guidelines and request form redesign		+	+	NR	−
Appropriateness (compliance)	Dowling et al26	Education and feedback		+	+	+	−	Per cent TSH indicated
	Nightingale et al32	Education and feedback and protocol management system		+	+	NR	−	Per cent patients requiring a particular investigation according to protocol who were actually tested
	Rhyne and Gehlbach43	Education and guidelines		+	−	−	−	Per cent ‘high’ and ‘low’ indications
	Schectman et al33	Feedback and reminders		+	−	−	−
	Toubert et al37	Guidelines and reminders		+	+	+	−	Per cent appropriate
Expenditure	Hardwick et al44	Funding policy and guidelines		+	+	NR	−
	Stuart et al34	Education and feedback and guidelines		+	+	+	−
	Tomlin et al36	Education and feedback and guidelines		+	−	NR	−
Pattern	Hardwick et al44	Funding policy and guidelines		+	−	NR	−	Sought to decrease proportion of TT3 as TFTs requested. Summary based on per cent of TFTs TT3
	Tomlin et al36	Education and feedback and guidelines		+	+	+	−	Sought to shift to TSH. Summary based on per cent TFTs TSH alone
	Toubert et al37	Guidelines and reminders		+	+	NR	−	Sought to shift to TSH. Summary based on per cent TFTs TSH alone
	van Gend et al38	Feedback and test form redesign		+	+	NR	−	Sought to shift away from TT4. Summary based on FT4:TSH ratio
	Vidal-Trecan et al40	Education and guidelines and request form redesign		+	−	NR	−	Sought to shift to TSH. Summary based on the proportion of FT3 and TSH
	Wong et al42	Guidelines and request form redesign		+	+	NR	−	Sought to decrease ordering of ‘complete’ thyroid panel to more selective use of individual tests

Direction: ‘+’ indicates result favours behaviour change intervention; ‘−’ opposite. Large effect: ‘+’ indicates risk difference is ≥20%; ‘−’ <20%. Statistically significant change: ‘+’ indicates 95% CI do not include no effect or p<0.05; ‘−’ opposite. RCT design: ‘+’ indicates study design is RCT; ‘−’ not RCT.

CV, coefficient of variation; ED, emergency department; FT3, free tri-iodothyronine; FT4, free thyroxine; GP, general practitioner; NR, indicates item not reported; RCT, randomised controlled trial; T3RU, tri-iodothyronine resin uptake; TFTs, thyroid function tests; TSH, thyroid-stimulating hormone; TT3, total tri-iodothyronine; TT4, total thyroxin.