Quality and performance improvement in critical care

In the past decade, there is an increased focus on quality and safety in health care. Decreasing variation, increasing adherence to evidence based guidelines, monitoring processes, and measuring outcomes are critical for improving quality of care. Intensivists have broad knowledge of hospital organization, and need to be leaders in quality improvement efforts.

[11] Parikh et al, evaluated quality of care at a public hospital in Mumbai, India, and reported a higher than expected mortality which may be related to multiple deÞ ciencies in delivery of care.In addition, the increase in travel tourism for health care to India is increasing, and there is a need to demonstrate outcomes comparable to other countries to compete effectively for this market.Public trust in health care providers could also be adversely affected if the public perceives that the care provided is not of high quality.The Medical Council of India (MCI) and the ministry of health are creating standards of care for physicians. [12]For these reasons, ICU performance need to be scrutinized closely to evaluate both the effectiveness of ICU treatments and the quality of care delivered in ICU.
The following review includes • History of Quality Improvement • Quality Improvement Methods and implementation • Quality Improvement Initiatives in ICU

Introduction
In the past two decades improvements in life-sustaining technologies (LST) resulted in an increase in the number of intensive care units (ICUs), and patient receiving LST in the ICUs.Care of the critically ill patients is resourceintensive, and 15-20% of hospital budgets are spent in the ICUs.[4][5] Poor quality care is not only costly but also causes human suffering because poor quality care results in increase in morbidity and mortality.Quality Improvement (QI) initiatives in the ICU to decrease nosocomial infections and maintenance of normoglycemia have been shown to improve outcomes as well as decrease costs. [6,7]Clemmer reported that improvement in quality of care in the ICUs at a tertiary care center resulted in an estimated savings of $2.6 million per year. [8]During the past decade, in India, of health care is very old as indicated by the admonition "Þ rst do no harm".Florence Nightingale kept records of her patients and outcomes to assess the impact of care, and suggested that knowledge of outcomes is crucial to improving care.Codman, one of the pioneers in QI in the early 20 th century, reported his outcomes in surgical patients and advocated pubic reporting of outcomes by both physicians and hospitals. [1]The modern QI initiatives started with recognition in other industries that unexplained variation leads to poor quality, and processes that decrease variation and continuous evaluation leads to improved quality.Shewhart and Deming were proponents of continuous evaluation of processes to improve quality and decrease defects. [13]onabedian [14] initiated the structure, process, outcome paradigm to improve health care, and Berwick and others applied these principles to the health care and led efforts to improve quality of care in the United States of America (USA) The report "To Err is Human" by the Institute of Medicine (IOM) in the United States in 1999, led to an increased focus on safety and quality of care.IOM suggested that care should be safe, effective, patient centered, timely, efÞ cient, and equitable.IOM reported that that one of the primary quality problems is inappropriate use of resources, and suggested efforts to improve the use of resources by focusing on overuse, under use and misuse. [15,16]As a response to the IOM reports, many institutions initiated QI efforts to improve quality of care.

Quality Improvement Methods and Implementation
Quality is deÞ ned by the IOM as "the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge."Although there has been more emphasis on performances of healthcare providers and quality of care recently, the focus on quality of care is not new.Codman suggested, about 100 years ago, that hospitals must collect data on their outcomes, identify strong and weak points and compare the results with other hospitals.Unexplained variation in patient care is based more on physician biases rather than patientrelated factors.In the study of intensive care physicians at a university hospital, Garland et al reported a 43% variation in resource use and costs (≈ $1,000) between intensivists without a signiÞ cant difference in mortality or length of stay. [17]Variation in care delivery makes it very difÞ cult to monitor processes and outcomes.
Although physicians accept improving quality of care as a goal, they are sometimes skeptical of quality improvement efforts and consider participation in QI efforts as a non productive use of their time and view efforts to decrease variation as an interference with their autonomy.Physicians' behavior is influenced by suggestions from a respected colleague or role model, appropriate support for professional skill development, reinforcement by colleagues, feedback from patients, and visible results.Physicians need to agree that processes that inß uence clinical activity lead to measurable outcomes.Feedback and reÞ ning the process based on clinician input would get buy-in from frontline staff.[20][21] Physicians attempting to lead QI efforts need to be cautious on how they interact with other physicians because a wrong approach could lead to failure although the intervention is effective.The story of Ignac Sammelweis, who was a pioneer on hand hygiene but was unable to inß uence his colleagues, illustrates that the person who wants to initiate change needs to be able to communicate his ideas to both his superiors as well as coworkers in a nonthreatening manner, and be cautious in how he conveys the message. [22]As intensive care physicians interact with many medical specialties and have a better knowledge of hospital organization because of their interaction both with the physicians as well as administrators, they are well suited to become leaders in QI initiatives.Donabedian proposed reviewing structure, process and outcome to improve quality of care. [14]The model is described in Figure 1.The Structure, in the ICU setting, refers to the type and size of the ICU, nature of stafÞ ng and availability of technology.Process issues include communication among staff, use of available technology and trainee guidelines and supervision.Outcomes include resource use, use of diagnostic and therapeutic procedures and mortality.Interventions affecting structure take longer to implement and are more expensive, so initially it is easier to target processes of care, modifying them as needed, and measuring the outcomes affected by the process.Some outcome measures such as length of stay (LOS), mortality, are easy to measure, but are affected by a number of variables and may not be easily attributed to a single intervention.
The success of QI projects depends on identifying projects which all stakeholders find useful and building a team culture.Performance measures and outcomes should be clearly defi ned, valid, and reliable.Documentation and data collection should be incorporated into daily work routines.Team development and process/ outcome defi nition, followed by an iterative process of implementation, evaluation and process adjustment based on the evaluation are important steps in achieving the goals.Leadership buy-in and support is essential for implementation and success.A comprehensive plan with a description of the goals, plan for implementation, cost and benefi ts with business plan and timeline will be helpful in obtaining administrative support.
Successful implementation of changes in practice are facilitated by check lists, disease specifi c pre-printed order sets, daily order sets that include goals for care.Standardized order sets facilitate implementation of best practices in addition to improving compliance with best practices.Establishing standards of care, monitoring processes and outcomes, creation of multidisciplinary teams, data recording as part of routine care, automated retrieval of information by using information technology facilitate QI efforts.A bedside electronic record facilitates data collection and retrieval.The experience of institutions with successful implementation strategies include: leadership support, incentives for senior leaders, physician and nursing leadership in implementing the initiatives, and involvement of bedside caregivers in the design and implementation of a QI projects.

Elements of Design and Implementation of QI Project
Process Identification of a clinical process that need be changed-based on benefits/risk/costs; patient care needs, informal discussions and payor priorities

Goals
Desired outcomes.

Measurement
Process measures, protocol, data

Reiterative process of Plan, do, Study, Act (PDSA) cycles
The improvement is usually incremental and requires repeated evaluations and refinement of processes.Figure 2 illustrates the PDSA cycle

Celebrate success QI projects in the ICU
A brief summary of the initiatives, categorized according to the principles of quality listed by IOM is provided in Table 1.Table 2     The process should involve multidisciplinary teams consisting of intensivists, ICU nursing staff and staff of respiratory therapy department with participation from other departments such as Infection Control and Blood Bank.Consensus guidelines which include inclusion and exclusion criteria, algorithms for implementing each of the process elements, deÞ nition of outcomes and data collection need to be created, As an example of one of the projects, the algorithm for implementation of daily Spontaneous Breathing Trials (SBT) was shown in Figure 3.The experience at University of Pittsburgh Medical Center (UPMC) with SBT indicated that although the compliance with daily SBT was high, the extubation rate is not optimal.So, we are evaluating the reasons for failure to extubate and will modify the guidelines and algorithms based on the experience.It has to be noted that the success of these projects requires sustained support from the administrative and medical leadership, a physician champion, and motivated team.As patients are heterogeneous in their diseases and acuity, co-morbidities and age, any evaluation of quality needs to consider these factors.It would be helpful to collect severity of illness information so that outcomes of patients in different ICUs could be compared but it adds to the costs of obtaining data.Risk adjustment models, such as Acute Physiology And Chronic Health Evaluation (APACHE) [23] or SimpliÞ ed Applied Physiology Score (SAPS), [24] adjust for these risk factors and allow comparison of different ICUs or, in some cases, evaluation of QI initiatives within a single ICU over time.
Ventilator Associated Pneumonia (VAP) bundle: VAP increase length of stay and morbidity. The components are listed below and the algorithm used at UPMC is shown in Figures 3-5 Sepsis Bundle: Standardized management of sepsis decreases costs and improves mortality. [27]Shorr reported that mortality was 20% lower, LOS was Þ ve days shorter, and costs were ≈ $ 5,000 lower in sepsis patients treated by protocol. [28]Early Goal Directed Therapy (ß uid resuscitation, vasopressor/ionotropic support) within six hours of identiÞ cation of sepsis

HOB Protocol
All intubated patients will have their head of bed (HOB) elevated to >30° unless specifi c contraindications exist or physician order specifi es supine position.Central Line Associated Bacteremia (CLAB) Bundles: Shannon et al reported that CLAB not only increases morbidity and but also resulted in a loss to the hospital because the reimbursement is lower than the costs.Implementation of the CLAB bundle resulted in a decrease of 825 (7.7 to 1.4 infections/1000 catheter days). [6]Pronovost et al reported that implementation of all elements of the bundle decrease CLAB rate from 7.7 to 1.4 infections/1000 catheter days. [29]e components of the bundle are  Rapid Response Team (RRT): Foraida et al and others reported that implementation of RRT response resulted in a decrease in cardio respiratory events leading to cardiac arrest and improved survival.RRT helps to identify patients at risk and provide early resuscitation. [31]he composition of RRT is variable but usually consists of an ICU nurse, respiratory therapist and a physician skilled in airway management.

Conclusion
The Institute of Medicine in the U.S. reported that there is a quality chasm in healthcare and suggested that the delivery of healthcare should be improved, so that it is safe, effective, patient-centered, timely, efÞ cient and equitable.Both medical leadership and staff need to work together to achieve such a healthcare system.Effective implementation of existing treatments that were shown to be beneÞ cial is more cost effective than implementing newer treatments that are marginally more effective.Intensivists, because of their broad knowledge of the hospital and interactions with multiple specialties are well suited for leading efforts to improve quality of care.

Figure 2 :
Figure 2: QI Process . • Head of the Bed elevation (HOB) to > 30 degrees • Daily Sedation Interruption (SI) • Daily spontaneous Breathing Trials (SBT) • Oral care • Deep venous Thrombosis (DVT) prophylaxis • Stress Ulcer Prophylaxis Outcomes: • Compliance with process measures (HOB, SBT, SI) • Ventilator days, length of stay (LOS) • Incidence of VAP -Pneumonias/1000 ventilator days • Reintubations Trendelenburg at 30° • Patients status post balloon pump or femoral arterial sheath removal for six hours • Post-surgical patients with open abdominal wounds and packing • Obese patients with a body habitus causing femoral lines, hemodialysis catheters or ECMO cannula to malfunction when the HOB is elevated • Liver disease patients with severe orthodeoxia (SpO2 < 90% when changing from the recumbent position to upright position) • Pancreas transplant recipient for three days post-op The following patients have both contraindications for HOB elevation and reverse Trendelenburg • Patients with hemodynamic instability upon elevating the HOB (> 10 mmHg drop in systolic blood pressure) • Neurological injury or Þ ndings referable to the thoracic or lumbar spine • Symptoms of back pain or neurological injury referable to the thoracic or lumbar spine • Radiographically identiÞ ed unstable fracture or T/L spine • Patients with spinal cord injury • Patients with unstable thoracic or lumbar fractures unless cleared by orthopedic spine surgery or neurosurgery • Open chest

Figure 3 :
Figure 3: Head of bed protocol ventilation and daily • Spontaneous breathing trials, glycemic control: location, duration • and review for catheters/tubes • Review of medications including antibiotic coverage • Pain/sedation management and daily sedation interruption • Nutrition, stress ulcer and DVT prophylaxis • Activity • Communication -with consultants, family Another tool to improve communication is to standardize format of communication between staff and physicians by following the SBAR tool • S-Situation: description of clinical situation • B-Background: clinical history/context • A-Assessment: a description of possible problems • R-Recommendations: a description of possible solutions contains further details of these initiatives.

Medication Recommendations Sedation -Propofol (maximum dose 80 mcg/kg/min) Delirium-Haldol Anxiety -Benzodiazipines Pain -Opiates Documentation 1. Prior to start of weaning , at 15 min. and 1 hr after start, and prn record HR, RR, BP, SpO 2 , and Ramsey score 2. In nurses notes summarize how pt tolerated sedation interruption
The criteria for calling at RRT at University of Pittsburgh Medical Center (UPMC) are listed below Respiratory • Rate <8 or 36/ minute • New onset difÞ culty breathing • New pulse oximetry reading <85% for > 5 minutes in a patient with no prior history of chronic hypoxia • New requirement for FiO 2 >0.5 to obtain SaO 2 >85% Other QI measures that could be evaluated in the ICU include: • End-of-life care and family support • Management of acute lung injury • Enteral nutritional support • Glycemic control in critically ill patients.