Table 1

Use of Theoretical Domains Framework (TDF): implications of the classification of partner types for partner notification (PN) approaches

Partner type*	Established partner	New partner	Occasional partner	One-off partner
Selected Theoretical Domains Framework functions that discriminate index patient’s engagement in PN	Knowledge
	Good knowledge of partner likely. This might facilitate PN; the index patient may anticipate their reaction and their respective choice of PN approach.	Uncertain knowledge of the partner, their reactions, their choice of PN approach.	Good knowledge of partner likely. This might facilitate PN; the index patient may anticipate their reaction and their respective choice of PN approach.	Very little of the partner may facilitate or constrain engagement in PN.
	Social role and identity
	(Eg, spousal roles) May facilitate motivation to engage in PN according to scripts, expectations and assumptions.	(Eg, romantic roles) May constrain engagement in PN, although novelty of relationship ‘permits’ residual infections.	Scripts and norms (eg, ‘friends with benefits’) may enable engagement in PN.	Identities (eg, being a player) may constrain engagement in PN.
Implications for PN	Beliefs about the consequences of engaging in PN are important and varied (health, interpersonal and infidelity).	Beliefs about the consequences of engaging in PN are important and varied (health, interpersonal and end of the relationship).	Beliefs about the consequences of engaging in PN are probably health oriented and may facilitate PN.	Likely to have minimal concerns about the consequences of engaging in PN.
	Emotions may be particularly important given length of relationship and likely expectations of exclusivity.	Emotions may be particularly important given potential emerging expectations of exclusivity; they may be particularly intense.	Emotions may be less important and not represent barriers to engaging in PN.	Emotions are not likely to be important in relation to engaging in PN.
	Highly amenable to PN interventions that use the existing relationship dynamics, rapid effective access to partners, established routes of communication and almost guaranteed future interactions between partners.	PN interventions that rely on existing relationship dynamics may be problematic and access to partners may be difficult.	Highly amenable to PN interventions that can use the existing relationship dynamics, established means of access to partners, established routes of communication and almost guaranteed future interactions between partners.	PN for these types of partner maybe more resource intensive as index patients may have low motivation to engage in PN.
	Depending on the nature of the particular relationship, there may be emotional issues that may delay or prolong PN. Additional support and services may be warranted.	Depending on the nature of the particular relationship, there may be emotional issues that delay, or constrain, some approaches to PN. These may require additional support and services.	There may be high motivation to engage in PN, and relatively few resources may be needed.	Patient referral may be challenging and may miss important public health opportunities.
		Framing effective PN as a commitment to the new relationship may be a beneficial approach.	PN approaches that motivate people to engage in PN because of health consequences and to protect their partner and themselves from future harms may be effective.	PN approaches that seek to motivate index patients through appeals to social norms and ideas of the social good may be effective. PN approaches that use the likelihood of reinfection and the health consequences to self are unlikely to work.
	PN approaches that rely on interpersonal dynamics (including accelerated partner therapy (APT)/expedited partner therapy (EPT)) are likely to work, although emotional aspects may be key barriers.	Approaches such as APT or wider EPT may be appropriate for some of these partners.	Approaches such as APT or wider EPT may be highly appropriate for these partners.	PN interventions that rely on interpersonal dynamics (including APT/EPT) are unlikely to work.
				May be particularly amenable to enhanced provider referral that engages fully with the mechanisms by which these partners met (eg, dating apps).

*The sex worker partner type was added to the classification after completion of this phase of work and so was not included in the TDF analysis.