Improvement cycles are critical to continued use of effective innovations and are an essential part of using effective Active Implementation methods.  To use innovations and implementation drivers in practice is to invite problems.  There are big problems such as changes in the economy or social expectations and there are local problems like working conditions, salary expectations, staff turnover, leadership changes, and so on that impact the use of innovations.  Any and all of these impact on-going human service delivery.  These environmental factors are coupled with the need to continually improve the usability and effectiveness of an innovation and of the competency, organization, and leadership drivers to produce and scale socially significant outcomes.

Implementation and improvement are conjoined twins joined at the heart.  To “choose one” is folly.  Improvement science needs implementation science.  Improvement methods themselves must be used as intended (with fidelity) if they are to be effective in practice.  Taylor et al. (2014) conducted a review of 73 published studies that reported application of plan, do, study, act, cycle (PDSAC) methods.  They found that only 2 (3%) of the 73 studies included all five components of PDSAC.  While all 73 had a plan there was virtually no indication of how well the plan was done, only 7 studied results in a systematic way, 2 indicated how they acted on the information/data, and 14 reported conducting an ensuing cycle.  Thus, if improvement methods are to be used fully and effectively in practice, then implementation methods must be used to assure their use as intended (with fidelity).  This point has been made by others who have studied the use of improvement methods in practice (Khodyakov et al., 2014; Leis & Shojania, 2017; Marshall et al., 2017; Øvretveit et al., 2017; Oyeledun et al., 2017; Reger, Gustafson, DeMarie, & Mullane, 1994; Shojania & Grimshaw, 2005; Westphal, Gulati, & Shortell, 1997).

Implementation science needs improvement science.  No matter how successful attempts are to achieve socially significant outcomes, there always is room for improvement.  There are no perfect innovations, totally effective implementation methods, or completely enabling contexts.  Implementation is a world of imperfection that is constantly striving for noticeable improvement.  Improvement cycles are based on trial and learning approaches to problem solving in all aspects of implementation.