A Great Debate in Implementation Science: Fidelity vs. Adaptation

The relative merits of fidelity and adaptation when implementing programmes have been discussed for decades. Even within more recent Implementation Science literature, there is an obvious tension between the two. Here, Chris Minch, weighs up the arguments for both sides and proposes that elements of both can be combined to maximise a programme’s potential for success.

Fidelity means implementing a programme in the exact way its developer initially designed it to be implemented. Key elements of this process include delivering components in the right order, complying with training procedures, and faithfully replicating all elements of the programme. This is important, particularly if the programme has been subjected to rigorous evaluation, because it means staying true to the evidence base.

Adaptation on the other hand involves changing certain elements of a programme or policy to make them potentially more appropriate to a specific context. Adapters reason that no two contexts can ever be the same (ranging from cultural differences across continents to fire alarms interrupting a training session) and perfect fidelity is virtually impossible in any real world situation. This gives them a mandate to adapt programmes to account for contextual differences.

Brick 101/Creative Commons License

Brick 101/Creative Commons License

So where does the tension between the two concepts arise? On one side of the ideological divide are a group I’ll call the ‘naysayers’ – programme developers who vehemently oppose any adaptation, seeing it as a failure of implementation, and tightly controlling how their programmes are implemented. On the other side are two different types of adapters. One – the (possibly reluctant) ‘accepters’ – see adaptation as a necessary element of implementation in the real world that can help improve traditional approaches around the margins. Alternatively, adaptation ‘drivers’ champion needs-led changes as central to positive programme development.

There are valid arguments against gung-ho adaptation. The main one put forward is that adaptation can lead to ‘programme drift’, whereby changes made to the programme during implementation are assumed to lower the overall efficacy of evidence based interventions. Another is that adaptation may make programme evaluation against the evidence base trickier because you are not necessarily comparing like for like.

Often, however, a key assumption is made that fidelity and adaptation are dichotomous concepts, both of which have a linear relationship with programme success. An alternative and, in my opinion, very reasonable viewpoint is that this assumption doesn’t hold in the real world. There is no convincing reason for why programmes couldn’t contain a degree of flexibility while still maintaining their original integrity and intent.

To highlight this, I’m borrowing an example from Harvard’s Matt Andrews. He describes the development of internal audit reform in Burkina Faso in 2007. Following a corruption crisis, the president brought in an academic, Henri Bassin, to implement a system for identifying financial risk. Bassin approached cabinet members to consider undergoing risk mapping and audit exercises. Five ministers agreed, forming the basis of an initial pilot, which used relatively simple auditing and risk mapping techniques with a high degree of fidelity. Within months, the project delivered widely publicised positive results.

Interest grew and the next pilot began to show signs of adaptation: it was expanded to 11 ministries and the auditing team were trained in more complex, performance-based techniques. Again, the pilot was a success and the practices of internal auditing became a hugely important part of the government’s work – so much so that even after a coup d’état, the agency was given an ever-expanding anti-corruption mandate, becoming the “jewel in the crown” of the new government.

John Taylor/Creative Commons License

John Taylor/Creative Commons License

The example contains a model balance of initial fidelity followed by adaptation, creating a dynamic audit programme that improved with time. Dynamism of this kind has been given various names, for example Problem-Driven Iterative Adaptation or Dynamic Sustainability.

Regardless of name, the most important takeaway is that fidelity and adaptation aren’t seen as antagonistic, ‘us vs. them’ concepts, creating conflict between programme developers seeking fidelity and implementers seeking adaptation. Instead, both can be combined to maximise a programme’s chances for success. For example, developers have theoretical and technical expertise which can help them to establish ‘core components’ of programmes to which changes should be minimised. Practitioners and implementers can then utilise their contextual knowledge to put these core components to best use.

Finding the ideal balance between fidelity and adaptation may prove difficult and only emerge through trial and error. However, the potential benefits mean that a deliberate effort must made to effectively combine both fidelity and adaptation. As Thomas Backer puts it: “Asking the question ‘is adaptation desirable?’ is asking the wrong question. Adaptation will happen, so the questions to be asked instead are: ‘how much?’ and ‘when is a programme’s content damaged?’”.


The views expressed in this article are those of the author and do not necessarily represent the views of Development in Action.

***

Have an opinion on this or another topic? Why not write for our blog? Click here to find out more and get in touch.