Loss to follow-up is a potential problem with any clinical trial. Some patients would have been lost to follow-up, but this would only change the results if nurses' patients were more likely to be lost to follow-up - usually, the number lost is similar between nurse and physician arms and, according to the reviewers, most trials had follow-up >80%.
Does the review have limitations? Yes, most notably with the short duration of follow-up. But I glad that policymakers are using this rather than some anecdotes about NPs harming patients or wasting resources.