Interventional cardiology consists of several related procedures that are performed in the coronary, peripheral, and cerebral vascular systems, as well as the central aorta, the cardiac valves, and the structural units (parenchyma) of the heart itself. Almost all of the procedures are performed under radiographic fluoroscopic guidance in a cardiac catheterization laboratory (cath lab), or sometimes a “hybrid” laboratory that can also function as a surgical operating room. Often the fluoroscopic imaging is complemented with intravascular ultrasound, transthoracic or intracardiac echocardiography, and rotational computed tomographic angiography (CTA). Thus, the modern cath lab is a complex, highly technologically sophisticated facility where both patients with chronic, stable conditions as well as patients with life-threatening illnesses are evaluated and treated. Therefore, it is essential to have an active quality assurance and improvement (QA/QI) program in place. This program will need to consider all aspects of the risks encountered by patients undergoing procedures in the cath lab, as well as by staff working there. This will of course include radiation risk and methodologies for reducing it. However, radiation risk is an extensive subject on its own and will not be covered in this chapter.

With the tremendous growth and increased experience of invasive cardiovascular training programs, paralleling the general overall growth in cardiac catheterization as a common diagnostic procedure, there has been a decline in the risks of undergoing an invasive procedure. Complication rates with general diagnostic catheterization and angiography are quite low. Major adverse cardiovascular events (MACE) such as death, stroke, myocardial infarction (MI), and emergency surgery typically occur in <0.1% of diagnostic procedures.¹ However, for cardiovascular procedures where therapeutic intervention is attempted, the story is very different (Tables 73-1 and 73-2). Complication rates may be much higher and may be widely different between institutions.

In addition to the MACE events listed in Tables 73-1 and 73-2, there are several other serious events that are of clinical importance and typically are incorporated into the QA/QI program. The foremost of these are acute kidney injury (AKI), bleeding, and blood transfusions. The reasons for this are illustrative of many of the larger quality issues in general.

Acute Kidney Injury (AKI)

A large-scale analysis of almost 1 million percutaneous coronary intervention (PCI) patients from >1200 hospitals revealed that AKI occurred in approximately 7.1% of the patients, and 0.3% (3000 patients) had a new requirement for dialysis.² A number of factors were found to be independently associated with developing AKI, and these could help serve as clinical predictors of the event (Fig. 73-1). The occurrence of AKI was quite serious, inasmuch as rates of death, bleeding, and MI were all significantly increased when AKI occurred (Fig. 73-2).

Bleeding and Transfusion

Invasive cardiovascular procedures can be accompanied by unwanted and unexpected bleeding. While it might be thought that blood transfusion is simple and effective for treating blood loss when necessary, many studies have shown that both bleeding and blood transfusions are serious clinical events in themselves and are harbingers of worsened outcomes. In a large analysis of >1 million PCI procedures at >1100 hospitals, major bleeding within 72 hours of the procedure was found in 5.8% of patients (60,000 cases), which is a very substantial number.³ Periprocedural bleeding has been found to be related to increased mortality by almost double over the subsequent year at least (Table 73-3 and Fig. 73-3).^4,5 Furthermore, analyses of blood transfusion events in PCI patients, even when adjusted for coexistent anemia and blood loss, have found transfusion to be related to long-term mortality (Fig. 73-4).⁶

To help identify ahead of time and possibly prevent bleeding in PCI, several bleeding risk models have been developed, one of which is shown in Table 73-4.³ This model is based on a risk scoring system and uses 10 clinical variables available at the bedside. Points are given for each level of the variables, and the summed points can be used to classify patients into low-, medium-, and high-risk groups for bleeding. The model has been found to be useful in identifying PCI patients at higher risk for bleeding events (Fig. 73-5).

Risk models such as these for AKI and bleeding, as well as others for risk-adjusted mortality, often are the primary focuses of a QA/QI program. This may especially be true in the early stages of a new QA/QI program that has just been initiated.

The quality imperative throughout medicine and cardiology, but particularly in procedure-based interventional cardiology, specifies that regular, comprehensive, and detailed analyses of (1) patient selection, (2) processes of care, and (3) outcomes will lead to reductions in adverse events and thereby to improvements in safety and quality of care. It is for these reasons that QA/QI programs in interventional cardiology have come into existence. One of the major focus areas for these programs is the variations in practices between operators and between institutions, and the possible relationships between such variations and adverse events. Variations therefore have become a major subject area for review. Examining variations is one way to begin to understand processes that may be at work and how they might relate to outcomes, both good and bad. Several examples can be given for illustration.

Detection of Coronary Artery Disease (CAD)

An analysis of >560,000 patients with no history of prior MI or revascularization, undergoing elective diagnostic catheterization at 691 hospitals found there was wide variation between hospitals in detecting obstructive CAD, even when allowances were made for the amount of obstructive disease that might be considered significant (Fig. 73-6).⁷ Variations in clinical practice patterns that determine suitability of patients for diagnostic angiography presumably underlie these variations.