By Ken Coburn, MD, DrPH, FACP, CEO and Medical Director at HQP
What’s most important? Replicating a proven care model with high fidelity in order to get reproducible results? Or adapting (i.e., changing) a proven care model to fit and work effectively within a specific context or environment? At first blush these two choices seem at odds, diametrically opposed to one another – increase one and you inevitably decrease the other. The cognitively uncomfortable truth is that, in most real-world settings of health care delivery, both high-fidelity replication and local adaptation are critical to the success of spreading even a “proven” model of care delivery to other sites. The scaling of a model that worked at original site #1 can fail by holding too fast to either concept at the expense of the other as one attempts to implement in sites #2, 3, etc.
It may be better to think of combining the two concepts (replication, adaptation) within a single optimization function through which one seeks to find a balance point that maximizes the probability that a new implementation of a proven model will be effective in achieving the desired health outcomes for a given setting and population. This may be especially relevant for models of care serving complex populations that require a broad set of dimensions to be effective; protocols, standards, and specifications, along with a compatible organizational or team culture, and principle-driven decision-making.
To innovate systems of replication that more effectively and efficiently spread and scale models of care, we must simultaneously innovate systems of adaptation and have the two work in synchrony. These must occur together since, in practice, they proceed very nearly at the same time and in the same place. Two systems? Or one? Some might argue that a new generation of “replication” systems could simply incorporate the concept of adaptation within them – replication now needing to address not only fidelity, but also adaptation. Perhaps. But this would likely lead to a linguistic and cognitive bias in which the classical meanings associated with replication are automatically assigned more importance than the newly engrafted concept of adaptation.
Instead, we could conceive of a single system of replidaption that facilitates both replication and adaptation simultaneously in a manner that sees both as critical variables in an optimization function – not in opposition to one another – but interrelated, with each contributing to the success of sites adopting and implementing a new model of care.
If there is utility in the term replidaption, it is not in the cleverness of its word play. A portmanteau of “replication” and “adaptation”, the power of the term is in changing the way we think and innovate – to one that is more integrated, holistic, and effective. To be useful, this way of thinking must ultimately offer some advantage to those deciding to design systems of replidaption rather than systems of spread that continue to struggle to reconcile the duality of replication and adaptation as separate and competing processes.
I think replidaption could prove to be a very useful term and mental construct because it is more likely to compel model innovators to think about how to apply their knowledge to optimize adaptation at each unit of spread; not simply whether or how much adaption to allow, but specifically how to adapt in ways most likely to preserve model effectiveness. At the same time, the construct calls upon adopters to play a more active role in providing an understanding of site context and the necessary inputs for adaptation while simultaneously stretching to achieve fidelity. I and the team at HQP are excited to collaborate with others, to further develop the concept and methods of replidaption and to design systems of replidaption that can be put to the test.
What’s most important? Replicating a proven care model with high fidelity in order to get reproducible results? Or adapting (i.e., changing) a proven care model to fit and work effectively within a specific context or environment? At first blush these two choices seem at odds, diametrically opposed to one another – increase one and you inevitably decrease the other. The cognitively uncomfortable truth is that, in most real-world settings of health care delivery, both high-fidelity replication and local adaptation are critical to the success of spreading even a “proven” model of care delivery to other sites. The scaling of a model that worked at original site #1 can fail by holding too fast to either concept at the expense of the other as one attempts to implement in sites #2, 3, etc.
It may be better to think of combining the two concepts (replication, adaptation) within a single optimization function through which one seeks to find a balance point that maximizes the probability that a new implementation of a proven model will be effective in achieving the desired health outcomes for a given setting and population. This may be especially relevant for models of care serving complex populations that require a broad set of dimensions to be effective; protocols, standards, and specifications, along with a compatible organizational or team culture, and principle-driven decision-making.
To innovate systems of replication that more effectively and efficiently spread and scale models of care, we must simultaneously innovate systems of adaptation and have the two work in synchrony. These must occur together since, in practice, they proceed very nearly at the same time and in the same place. Two systems? Or one? Some might argue that a new generation of “replication” systems could simply incorporate the concept of adaptation within them – replication now needing to address not only fidelity, but also adaptation. Perhaps. But this would likely lead to a linguistic and cognitive bias in which the classical meanings associated with replication are automatically assigned more importance than the newly engrafted concept of adaptation.
Instead, we could conceive of a single system of replidaption that facilitates both replication and adaptation simultaneously in a manner that sees both as critical variables in an optimization function – not in opposition to one another – but interrelated, with each contributing to the success of sites adopting and implementing a new model of care.
If there is utility in the term replidaption, it is not in the cleverness of its word play. A portmanteau of “replication” and “adaptation”, the power of the term is in changing the way we think and innovate – to one that is more integrated, holistic, and effective. To be useful, this way of thinking must ultimately offer some advantage to those deciding to design systems of replidaption rather than systems of spread that continue to struggle to reconcile the duality of replication and adaptation as separate and competing processes.
I think replidaption could prove to be a very useful term and mental construct because it is more likely to compel model innovators to think about how to apply their knowledge to optimize adaptation at each unit of spread; not simply whether or how much adaption to allow, but specifically how to adapt in ways most likely to preserve model effectiveness. At the same time, the construct calls upon adopters to play a more active role in providing an understanding of site context and the necessary inputs for adaptation while simultaneously stretching to achieve fidelity. I and the team at HQP are excited to collaborate with others, to further develop the concept and methods of replidaption and to design systems of replidaption that can be put to the test.