Why Better Control Is Not the Same as Fewer Events
Pharmacist-integrated collaborative outpatient care is the subject of this series: a model in which pharmacists work with primary-care teams and patients on medication management, follow-up, adherence, education, and monitoring. The question is not simply whether adding a pharmacist sounds useful. It is what this coordinated model changes, how reliably it changes measurable risk factors, and whether those measurements are enough to support a claim about fewer cardiovascular events. The answer matters because cardiovascular-kidney-metabolic (CKM) risk accumulates over years, while clinical care often records progress through nearer-term measures such as blood pressure and glycated hemoglobin (HbA1c).
The central judgment is deliberately narrower than a slogan. The available randomized evidence makes a strong case that pharmacist-integrated collaborative care can improve clinic systolic blood pressure. It offers a favorable but less stable signal for HbA1c. It does not, on its own, establish that the model prevents heart attacks, strokes, heart-failure admissions, kidney events, or deaths. Understanding why those statements can all be true at once is the purpose of this series.
The CKM problem begins before the event
The CKM framework treats obesity, insulin resistance, type 2 diabetes, chronic kidney disease, and cardiovascular disease as connected parts of a biological continuum rather than as unrelated diagnoses visited in separate rooms. The connections are not merely administrative. Metabolic dysfunction and excess adiposity can contribute to inflammation, oxidative stress, endothelial injury, neurohormonal activation, and progressive damage to target organs. In time, those processes can appear as cardiovascular or renal disease. The framework therefore shifts attention upstream: what can outpatient care do while risk is accumulating, before the most consequential outcomes occur?
That question is difficult because the outpatient setting is longitudinal but fragmented. A patient may have several interacting risks, multiple prescriptions, changing adherence, and limited opportunities for timely review. A treatment may be clinically appropriate yet not intensified when a target remains unmet. A medication plan may be sound in the clinic but difficult to follow at home. Measurements may be taken without becoming new decisions. The review identifies therapeutic inertia, delayed treatment intensification, medication nonadherence, and fragmented follow-up as reasons that routine risk-factor control remains suboptimal.
These are not gaps that a single professional can reliably solve by working harder in isolation. They involve information, authority, timing, and continuity. Someone must notice that a value remains above the desired range. Someone must review whether the regimen is suitable and whether barriers are practical, financial, or behavioral. Someone must decide what can change, communicate that decision clearly, and check what happened afterward. The primary-care provider, pharmacist, patient, and other team members may each hold part of that chain.
This is why the subject is a care model rather than a job title. A pharmacist's presence is not itself a mechanism. The relevant intervention is the organized relationship among medication expertise, clinical responsibility, patient experience, and repeated follow-up. In the randomized evidence reviewed by Zhao and colleagues, eligible approaches included recommendation-based collaborative medication management, protocol-based co-management, structured medication review, adherence support, patient education, and longitudinal follow-up. The models differed in how much authority pharmacists had and how intensely they were integrated into care, but they shared a disease-management function rather than merely dispensing medicines or offering one-time advice (Frontiers systematic review).
The distinction between a risk factor and an event is just as important. A clinic systolic blood-pressure reading is a measurement of a physiological state at a particular time and under particular conditions. HbA1c summarizes average blood glucose over a preceding period, but it is still a surrogate for later complications rather than a complication itself. A lower value may be valuable because it moves a person away from a risk state. Yet the measurement does not record whether a stroke was prevented. That requires a different outcome, a longer observation period, and enough participants for the event difference to be estimated credibly.
The same logic applies to target attainment. Counting how many people cross a prespecified threshold can make a result easier to visualize, but the threshold may differ between studies. The count is affected by baseline risk, follow-up timing, measurement error, and the number of people whose outcomes are actually observed. A target-attainment result can therefore be useful without being interchangeable with a continuous change—and neither automatically becomes evidence of fewer clinical events.
The distinction is not a technical disclaimer added after the fact. It determines what the evidence is allowed to say. A care model may improve a measurement through more timely treatment, better adherence, or closer monitoring. A separate body of evidence may show that sustained changes in that measurement are associated with lower event risk. Those two findings can form a plausible chain, but they are not the same trial result. The bridge must be tested, not assumed.
A care model is more than adding a pharmacist
The practical model begins with a shared care plan. CDC guidance describes team-based care as an intentional organization of a multidisciplinary team around the patient and the primary-care provider. Responsibilities can be distributed: one team member may support follow-up, another may review medicines, another may coach self-management, while the primary-care provider retains the overall clinical relationship and judgment. Depending on the agreement and local rules, a qualified pharmacist may recommend changes or manage medicines under a more protocol-driven arrangement. The important feature is not that every team uses the same division of labor. It is that the division is explicit enough for work to move rather than stop between roles (CDC, Team-Based Care to Improve Blood Pressure Control).
Consider the chain in a typical outpatient problem. A patient's home readings remain high, or a clinic measurement suggests that the current regimen is not achieving its intended effect. The team reviews the pattern, the medicines, possible adverse effects, missed doses, cost or access barriers, and the patient's goals. The pharmacist may identify an adjustment to discuss or implement within the agreed authority. The patient receives a treatment plan in language that is consistent across team members. Follow-up then tests not only the next number but also whether the plan was feasible. If the value remains uncontrolled, the new information prompts another review instead of waiting passively for the next routine visit.
Several mechanisms operate in that chain. More frequent therapeutic review can reduce the time between a missed target and a treatment decision. Adherence support can reveal that an apparently ineffective medicine is actually being taken irregularly, or that the prescribed plan conflicts with daily life. Home blood-pressure monitoring can add observations between appointments. Education and coaching can help a patient understand what to measure and why the measurement matters. Shared records and clear communication can reduce contradictory instructions. None of these steps is a magic ingredient, and the supplied review did not isolate each component as an independently proven cause. Together, however, they explain why the model could change an intermediate outcome.
The patient is not the final recipient of a decision made elsewhere. They are a source of information about symptoms, routines, access, priorities, and the actual use of medicines. CDC guidance recommends self-management plans, regular communication, appointment and treatment-plan confirmation, adherence support, and supported self-measured blood pressure. This matters because the care process extends beyond the clinic. A treatment plan that cannot survive ordinary life is not a completed intervention; it is an instruction awaiting adaptation.
The system can fail at each handoff. If nobody has authority to act, review becomes documentation. If team members send different messages, the patient must reconcile the plan alone. If follow-up is irregular, the team may not know whether a change worked or was never implemented. If a home monitor is supplied without training or clinician support, more data may create little action. If digital communication assumes reliable devices, transport, connectivity, or health literacy, the same model may reach some patients more easily than others. These are not arguments against collaboration. They are reasons to define the intervention precisely before asking whether it works.
The evidence base also contains variation that makes precision important. The review included 21 randomized trials across hypertension, type 2 diabetes, and mixed CKM-relevant populations, mostly in primary care but also in specialty and community ambulatory settings. The interventions ranged from moderate, recommendation-based collaboration to higher-intensity, protocol-based co-management. Some trials could be pooled for a particular outcome; others lacked comparable definitions, outcome-level denominators, or the variance data needed for quantitative synthesis. A result from one coherent subset should not be silently generalized to every version of the care model.
That point connects implementation to evidence interpretation. A health system cannot reproduce a pooled estimate simply by changing an organizational chart. It must ask who reviews the data, who can recommend or adjust therapy, how the team communicates, how often follow-up occurs, and how the patient participates. It must also account for labor, communication tools, staffing, reimbursement, training, and local barriers. CDC identifies role clarity, guideline-informed protocols, structured follow-up, patient engagement, technology, and community adaptation as implementation considerations. These choices affect whether the mechanism is present long enough to produce a measurable result.
The endpoint boundary is the point
The strongest result in the review is also where overstatement becomes most tempting. Three trials contributed adjusted estimates to the primary clinic systolic-blood-pressure analysis. Compared with usual care, pharmacist-integrated collaborative care was associated with a mean difference of −7.11 mm Hg, with a 95% confidence interval from −9.65 to −4.58 and observed heterogeneity of I² = 0.0%. The review describes this as its most robust and internally consistent quantitative finding. Sensitivity analyses supported its direction: no high-risk trial contributed to this pool, and omitting any one contributing study did not materially change the result.
That is meaningful evidence of an intermediate hemodynamic effect. It is not a small observation merely because it stops short of a heart attack or stroke. Blood pressure is a clinically important risk factor, and sustained improvement can matter. But the correct conclusion remains “the care model was associated with lower clinic systolic blood pressure in this randomized evidence,” not “the care model has been shown to prevent cardiovascular events.” The first statement follows the measured endpoint. The second requires direct event data or a carefully justified chain of evidence that this review did not supply at the same strength.
The difference becomes clearer when the other results are considered. Continuous HbA1c also favored pharmacist-integrated care, with a pooled mean difference of −1.26 percentage points, but the heterogeneity was high at I² = 89.2%. When the high-risk Javaid 2019 trial was excluded, the estimate attenuated to −0.81 percentage points and the 95% confidence interval ranged from −2.11 to 0.49. The glycemic signal is therefore favorable in direction but more sensitive to study quality and context. Blood-pressure target attainment produced a directionally favorable risk ratio of 1.44, yet its confidence interval was 0.68 to 3.04 and heterogeneity was 88.3%. HbA1c target attainment was based on only two studies and had a very wide interval, 0.08 to 31.35. These are reasons for calibrated confidence, not reasons to treat every outcome as equivalent.
This series will follow that distinction through the evidence and into implementation. It will examine why the blood-pressure pool is comparatively coherent, why the glycemic result is more uncertain, how target definitions and study quality shape interpretation, and what a functioning team must do between visits. It will then return to the question that governs the whole subject: what additional evidence is needed before better risk-factor control can responsibly be described as fewer cardiovascular or kidney events?
For now, the answer is both encouraging and bounded. Pharmacist-integrated collaborative outpatient care is a credible strategy for improving risk-factor management, particularly clinic systolic blood pressure, because it can organize medication expertise, follow-up, patient feedback, and timely action around a shared plan. Its promise should not be weakened by demanding that every intermediate result already prove mortality reduction. Nor should the promise be inflated into an event-prevention claim that the evidence has not yet tested. The value of the model—and the seriousness of evaluating it—begin with keeping those two judgments separate.