SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2398093498> ?p ?o ?g. }

Showing items 1 to 99 of 99 with 100 items per page.

W2398093498 endingPage "e41" @default.
W2398093498 startingPage "e33" @default.
W2398093498 abstract "After completing this article, readers should be able to: The issue of quality health care has achieved growing prominence over the last 2 decades. (1)(2)(3)(4)(5) The public reporting of “quality”; the use of quality as an important factor in negotiating care contracts; and activities to improve the quality of care at the practitioner, managed care organization, and hospital levels are now widespread and constitute what has been labeled the “quality revolution.” (6)(7)The Institute of Medicine (IOM) defines quality 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.” (8) This definition has been widely accepted because it speaks to a broad spectrum of stakeholders. The inclusion of individuals and populations addresses concerns that range from the clinical care of individual patients to population initiatives to improve socioeconomic and racial disparities in accessibility to and the quality of both preventive and curative health care. (9) The concept of desired health outcomes includes the desires of the patient to participate in and be satisfied with the medical decision making process as well as the desires of corporate and governmental payers to obtain value for their health dollars. (10) The inclusion of current professional knowledge emphasizes both the role of evidence-based medicine and the obligation for effective continuing education. Although the discussion of quality is as broad as the various stakeholders, this article addresses the topic of perinatal quality assessment from the perspective of the practitioner.In the very recent past, there was a general perception that the quality of medical care was beyond question, medical mishaps were unfortunate and uncommon events, local physicians and community hospitals provided solid routine medical care, and large urban medical centers offered unquestioned subspecialty expertise and cutting-edge equipment to care for complex or rare disorders. These perceptions shaped medical education, and issues of quality assessment were not addressed in either curricula or board exams. One of the first major challenges to this perception came in the late 1970s with a series of studies on the variability of illness, hospitalizations, and outcomes in differing geographic locations. (11)(12) The studies demonstrated that population illness rates did not explain population hospitalization rates, and even after controlling for differences in age and underlying health of patients, significant differences in the quality of outcomes achieved occurred in various hospitals. The demonstration that certain hospitals had consistently better outcomes than others brought the issue of quality to the fore.Table 1 shows the range of standardized neonatal mortality rates (NMRs) observed in California hospitals that had delivery services for infants born from 1995 to 2000. To increase sample size and achieve greater stability of the NMR estimate, a 5-year period was chosen. (The disadvantage to improving the stability of an estimate obtained from aggregating several years of data is that it obscures temporal trends in mortality that may have occurred during the time period. For this reason, it is appropriate to combine as few years as possible to obtain a stable estimate.) Even after controlling for differences in birthweight, race/ethnicity, and multiple births, marked variations in NMR were evident across California’s delivery hospitals. Using risk-adjusted NMR as an outcome indicator of quality of care provides strong evidence that efforts to improve perinatal care are warranted. (13)(14)Although efforts to uncover the causes of poor care and to develop strategies to improve care must be developed within the context of a specific situation, the IOM describes three important processes of concern: underuse, overuse, and misuse. (4) Examples of underused processes in perinatal medicine that have been shown to have significant health outcomes benefit include first trimester entry into prenatal care, the administration of antenatal steroids to mothers threatening preterm delivery, (15) hand washing to decrease the rate of nosocomial infection, (16) and the use of human milk in preterm infants. (17) Examples of overuse of activities at levels that have little evidence of benefit and actually may be associated with significant harm include excessive cesarean delivery (18) and excessive postnatal steroid use. (19) Both activities are essential when indicated, but excess use may be associated with operative complications and additional costs (cesarean delivery) and the risk of long-term brain insult (postnatal steroids), which may impose significant burdens on the patient. Misuse usually is described as medication errors, but it also could be extended to situations that are handled suboptimally, such as failure of communication between tertiary units and local caregivers or ineffective communication with parents. (20)(21)(22) Patient satisfaction, a quality indicator that is becoming increasingly important to payers, could reflect ineffective communication with parents. (10)An additional quality goal is optimal use. From the perspective of the clinician, optimal use is of paramount importance in motivating quality improvement activities. Whereas measuring underuse, overuse, and misuse depends upon observing processes (ie, gathering process measures of quality), measuring optimal use relies on analysis of outcomes. The concept of optimal use stems from observed variations in outcomes across similar populations. Such variation implies that some practitioners have achieved more optimal health care; that is, given a specific clinical situation, some combination of their clinical assessment, decision-making, therapeutic choices, and system for delivering therapy yields superior results. A further observation that stresses the possibility of optimizing care comes from a growing literature of evidenced-based medicine. It is now possible to assess the extent to which specific therapeutic approaches are effective. For example, randomized clinical trials and meta-analyses have been instrumental in demonstrating the efficacy of surfactant use and the relative effectiveness of early versus late administration of this therapy. (23)Quality assessment can be based on the examination of structures, processes, or outcomes. (2)Structural aspects of medical care include the availability of staff and services that are essential to meet patient needs. It is essential for compelling evidence to document that structural features affect outcome. For example, the percentage of very low-birthweight (VLBW) infants who are delivered in hospitals that do not have neonatal intensive care units (NICUs) is a valid structural indicator of perinatal regionalization because several studies have demonstrated that NMR is increased in VLBW infants who are delivered in hospitals that do not have NICUs. (24) It is important to note that although having no more than 10% VLBW infants in one region delivered at hospitals that do not have an NICU is a health objective for the nation, (25) the validity of this structural indicator is based on studies of differential mortality rather than the establishment of a national health objective by expert consensus.Another structural assessment that might be expected to vary across hospitals and settings is the ability to provide quality care regardless of the day of week or time of day. At the state level, NMR in California is not elevated for infants born on the weekend, although increased morbidity and mortality may be a possibility in hospitals that are severely understaffed on weekends. (26) Several studies from Europe have demonstrated a mortality disadvantage for infants born at night, suggesting a structural compromise in the quality of perinatal care. (27)A structural measure that has gained substantial notoriety is average NICU census. Although there is compelling evidence of improved outcome in certain adult surgical procedures performed at high-volume hospitals, the relationship between volume and outcome in the NICU remains unclear. One study states that NMR is decreased with volume; (28) another shows no volume effect. (29) Given the state of the evidence, it is unfortunate that one group of health care purchasers has adopted “an average NICU census greater than 15” as an NICU quality criterion. (30)Process indicators assess the use of specific processes in specific clinical situations. The assessment can be designed to identify underuse (eg, antenatal steroids) or overuse (eg, cesarean section). As with structural indicators, it is essential that compelling evidence suggest that the process is related to a positive or negative heath outcome. The dramatic increase in cesarean sections over the last several decades and the demonstration that many cesarean sections were performed in the absence of appropriate indications has led to a national initiative to decrease cesarean delivery. (25) The evidence appears to justify the use of excess cesarean rates as an indicator of suboptimal quality. Because high cesarean rates are unjustified and are valid indicators of overuse, it has been assumed widely that low rates of cesarean sections represent an indication of quality care. However, a recent study from the state of Washington clearly demonstrates that morbidity is increased in situations where cesarean deliveries are lower than expected. (18) The authors of the study emphasize the importance of validating the utility (or danger) of a process before using it as a quality indicator.The use of antenatal steroids to reduce the severity of respiratory distress syndrome (RDS) and intraventricular hemorrhage in mothers between 24 and 34 weeks of gestation who are threatening preterm delivery is in stark contrast to the example of low cesarean rates. The large body of evidence demonstrating the effectiveness of a single 7-day course of antenatal steroids was summarized in the 1994 National Institutes of Health Consensus Development Conference on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes. A second consensus conference was held in 2000 to evaluate the emerging practice of administering multiple courses. The second conference concluded that “because there is insufficient evidence from randomized clinical trials regarding efficacy and safety, repeat courses of corticosteroids should not be used routinely.” (15) From a quality assessment perspective, two processes had been validated as quality indicators: 1) Not administering a single course of corticosteroids to all women between 24 and 34 weeks’ gestation who are at risk of preterm delivery within 7 days (a measure of underuse), and 2) Routinely administering multiple courses of steroids (a measure of overuse).After validating that a process can have a significant impact on clinical outcomes, the first step in quality assessment using process indicators is to determine whether the practitioner (eg, physician, hospital, managed care organization, health department) is in a position to undertake the indicated action or therapy. Without this step, clinicians may be held accountable for not taking an action in situations where the action either was contraindicated by special circumstances or impossible to execute. For example, when calculating the percentage of preterm mothers who receive antenatal steroids, it is important that the denominator not include mothers who present too far advanced in active labor to qualify for steroid administration.Establishing accountability also is important when assessing quality using outcomes. An extreme example is a large public hospital that received public condemnation for having one of the highest fetal mortality rates in the state. On further investigation, the excess fetal mortality was shown to have occurred in women who had not received prenatal care at the facility and presented initially with fetal demise. Although high fetal mortality indicated the possibility of a serious quality of care problem (most likely in the domain of barriers to the use of prenatal care), it did not reflect the quality of intrapartum care provided by the hospital. Unfortunately, the hospital’s reputation was seriously compromised by the faulty assignment of performance accountability.Outcomes such as death, hospital length of stay, and various morbidities (eg, chronic lung disease, nosocomial infection) also can be used to assess quality of care. When assessing quality using outcomes, it is essential to consider the risks and comorbidities (often referred to as the case mix) of the subject patients. For example, when asked to choose between referring a patient to hospital A, which has an NMR of 1.6, or hospital B, which has an NMR of 2.7, it appears logical to choose the hospital that has the lower NMR. However, it is impossible to make this choice without also considering the risk and case mix characteristics of the mothers who deliver at the two hospitals (Table 2). In this example, Hospital A provides only primary perinatal services and on the basis of its case mix would be expected to have an NMR of 1.1. Hospital B has a community NICU. Because of its more complex case mix, it would be expected to have an NMR of 3.9. Hospital A has a higher NMR and Hospital B has a lower NMR than would be expected based on case mix. The difference between the two hospitals can be assessed more formally by calculating their standardized mortality ratios (SMRs). SMR is calculated by dividing observed NMR by expected NMR. The SMR of hospital A (1.6/1.1) is 1.46, indicating that its NMR is 46% higher than expected. The SMR of Hospital B (2.7/3.9) is .69, indicating that its NMR is 31% (1 − .69) lower than expected. From this analysis, it appears that hospital B provides “superior care” and hospital A provides “inferior care.”Even with the differences in case mix, before accepting this conclusion, it is essential to consider the role of chance. The role of chance can be appreciated in the paradigm that the observed rate of any outcome (and the observed rate for any process) depends on case mix, quality of care, and chance. (31) If the chance component is large, it may not be possible to conclude confidently that a high rate of adverse outcome represents poor quality. Two effective strategies for reducing the impact of chance are to increase the number of observations (usually accomplished by combining years) and to use quality indicators that have a high incidence of occurrence. (32)(33) To detect a 50% increase in an adverse outcome with the confidence that the observed increase is due to chance no more than 5% of the time (P<.05) requires 39,000 births for neonatal mortality (1.2 versus 1.8 deaths per 1,000 live births at P<.05, two-sided with a power of 0.8), 1,573 admissions for increased rates of nosocomial infection in infants weighing more than 1,500 g (3% versus 4.5%), and only 297 admissions for increased rates of nosocomial infection in VLBW infants (16% versus 24%). The details of the various statistical techniques commonly employed are beyond the scope of this presentation, (34) but all assessments of quality must be accompanied by a statistical estimate of the likelihood that what has been observed could have occurred due to chance.The estimated quality of care in primary care Hospital A and community NICU hospital B was determined by comparing the SMR based on case mix complexity. This approach allowed for adjustment in terms of differences in risks and comorbidities between the two patient populations. Risk adjustment has been described as a strategy for “leveling the playing field.” More formally, risk adjustment is “the process of sorting patients in each comparison group into different risk strata and then making comparisons separately for each stratum.” (33) Two general approaches are used. In the “single stratum” approach, a grouping of patients is defined with a specific set of risks for an indicator outcome, this group is identified in each of the settings being compared, and outcomes of these similar risk groups are compared directly across institutions. For example, NMR across NICUs could be compared for a single stratum defined as singleton VLBW infants who do not have birth defects. The advantage of this technique is that it is straightforward. The disadvantage is that the set of risks (plurality, birth- weight, and the presence of birth defects) may not include all of the important risk factors for mortality that could differ across institutions. In an analysis of VLBW infant mortality in the California Perinatal Quality Care Collaborative (CPQCC) network of NICUs in 2000, antenatal steroids, mode of birth, low Apgar score, gender, and intrauterine growth were statistically significant risk factors for NMR that had not been adjusted for by the previous single stratum definition. This obstacle could be overcome by enlarging the requirements of the single stratum, such as only considering singleton appropriate-for-gestational age males who had normal Apgar scores, no birth defects, and a specific birthweight and who were born to mothers who were given antenatal steroids.A more facile solution is to use an analytic approach that includes multiple risk factors by constructing a risk matrix or building a model based on regression equations (given the same risk factors, the results obtained from both strategies are expected to be similar). California’s Perinatal Profiles estimate expected hospital NMR for all delivery hospitals in California based on a risk matrix whose 138 cells are built from combinations of birthweight, race/ethnicity, gender, and plurality. The CPQCC estimates expected NICU mortality for infants whose birthweights are less than 1,500 g using a regression model that addresses inborn/outborn, birthweight, gestational age, intrauterine growth, gender, plurality, and 5-minute Apgar score. (35)When examining reports that present results adjusted for differences in risk or case mix, it is essential to assess the adequacy of the adjustment. This assessment requires two questions: 1) Have all of the factors known to exert a clinically significant risk of adverse outcome been included? and 2) Are the risk factors included in the adjustment completely out of the control of the entity being evaluated? Answering these questions requires experience. The first requires experience with the literature: What risk factors and comorbidities have been reported as significantly altering the likelihood of the index outcome and are they included in the adjustment? The second requires clinical experience: Are all of the risk factors in the model beyond the practitioner’s control? Table 3 lists factors that increase the risk of VLBW mortality, a common measure of NICU quality. When evaluating NICU quality across hospitals, it is appropriate to adjust for differences in mode of delivery and antenatal steroids because these decisions are not the direct responsibility of the neonatologist. In contrast, when evaluating VLBW mortality as an indicator of the quality of a hospital’s perinatal services (obstetric and neonatal), it is not appropriate to adjust for differences in cesarean delivery and antenatal steroid administration because these procedures are obstetric therapeutic interventions.In general, there are two broad classes of risk adjusters. An elevated risk of adverse outcome can be indicated by factors observed prior to neonatal intervention, such as birthweight, gender, race/ethnicity, and antenatal steroid use and by severity of illness scores based on physiologic features during the course of treatment. Historic (prior to medical intervention) factors are compelling in that many epidemiologic studies have validated an increased risk of adverse outcome on the basis of information available at the moment of birth. Severity of illness is especially compelling because clinicians accurately estimate mild illness in some neonates compared with the need for extreme levels of support in other infants. (33) Several severity of illness scores are available; the SNAP and CRIB recently have been recalibrated and updated. (36)(37) A potential problem in using these scores as risk adjusters is that they are based on observations made at 6 hours after birth, and physiologic derangements seen at this time could be influenced by the quality of care received during the first 6 postnatal hours. Furthermore, although historic data are collected routinely and usually can be obtained from the birth certificate or a computerized birth certificate file, the physiologic measures that form the core of the illness severity score must be abstracted from the medical record and entered into a database. Nonetheless, severity of illness remains an important risk adjuster. A key area in which severity of illness scores could be essential for risk adjustment is for neonates whose birthweights are greater than 2,500 g. For infants weighing less than 2,500 g, birthweight is a powerful predictor of outcome. A birthweight greater than 2,500 g is a weak predictor of adverse outcome, making severity of illness a very useful technique for risk assessment. In addition to their use as risk adjusters for quality assessment, severity of illness scores have been used as outcome measures to assess temporal changes in the quality of obstetric care and have been used to study the relationship between severity of illness and immune and hematologic function. (38)(39)Process and outcome each have advantages and disadvantages as indicators of quality. (40)(41) They both have “validity” requirements. Process indicators serve as valid quality indicators only when it can be demonstrated that variations in the process result in differences in outcome. Some advocate that outcomes should be used only as quality measures when there is evidence that specific variations in care result in differences in outcomes. However, benchmarking, surveys of outcomes across a network of NICUs to identify units that have very low levels of specific adverse outcomes, identification of features of care that appear to be unique, and use of these findings as potential improvement strategies also have supporters. (42)The first advantage of process measures is that they may be more sensitive quality measures than outcome data because poor outcomes do not occur with every error in the provision of care. For example, there are more instances of errors in drug orders (process) than in clinically recognized effects of drug errors (outcomes). A second advantage is that the evaluation of process is immediate; the unfolding of outcome may take much longer. Although the time factor is an important concern in adult medicine where failure to treat hypertension adequately could take decades to manifest as an increase in heart attack, the time course between process (such as failure to administer surfactant to a preterm infant who has RDS) and outcome (severe respiratory dysfunction) generally is very rapid during the perinatal period. An additional advantage of process is that because it is assessed in the context of a specific indication, there usually is little need to adjust for the presence or absence of comorbidities and risk factors. One defines the situation that calls for action, after eliminating those situations where certain factors that are out of the control of the practitioner make action impossible, and counts the rate of appropriate action.The advantage of measuring outcome as a quality measure is that it is extremely compelling. In a sense, outcome is “where the rubber meets the road.” A second important advantage is that outcomes may signal the possible effects of a variety of upstream structural changes (reduction in respiratory therapists, decrease in budget, use of nurses who do not have specific NICU training) even when the specific pathways by which structural issues compromise outcomes have not yet been identified. To be effective upstream signals, outcomes must be domain-specific and based on reasonable hypotheses. For example, fewer respiratory therapists could compromise the quality of respiratory care, leading to an increase in the average length of time on the ventilator. Fewer nursing personnel could bring about a higher workload, resulting in a breakdown in hand washing and increased nosocomial infection. A further advantage of using outcomes to evaluate quality is that outcomes, such as meconium aspiration syndrome (an indicator of the quality of postterm perinatal management), are captured routinely in both administrative and clinical databases. (43)(44)(45) In contrast, process measures, such as the extent of fetal monitoring in a mother beyond 42 weeks’ gestation (a process that is important in reducing the incidence of meconium aspiration syndrome) requires a specific chart review. Some processes (eg, antenatal steroid use or the time of initial surfactant administration) are so important that they now are captured routinely by a variety of clinical databases. (46)Ultimately, the choice of quality measure depends on local concerns and circumstances. Whether the measure is structure, process, or outcome, quality assessment always involves comparing observation with expectation. When structure, process, or outcome falls short of expectation, and the element of chance has been taken into account, improvement is required. Subsequent articles in this series in NeoReviews describe the process of quality improvement and illustrate how quality assessment is put into action to improve care." @default.
W2398093498 created "2016-06-24" @default.
W2398093498 creator A5007535386 @default.
W2398093498 date "2004-02-01" @default.
W2398093498 modified "2023-09-26" @default.
W2398093498 title "Quality Improvement in Perinatal Medicine" @default.
W2398093498 cites W1620660462 @default.
W2398093498 cites W1865720520 @default.
W2398093498 cites W1953921433 @default.
W2398093498 cites W1966304368 @default.
W2398093498 cites W1966940033 @default.
W2398093498 cites W1972486421 @default.
W2398093498 cites W1972584951 @default.
W2398093498 cites W1978036628 @default.
W2398093498 cites W2009312581 @default.
W2398093498 cites W2021487797 @default.
W2398093498 cites W2025243882 @default.
W2398093498 cites W2033817893 @default.
W2398093498 cites W2041648758 @default.
W2398093498 cites W2043072190 @default.
W2398093498 cites W2051829562 @default.
W2398093498 cites W2053339749 @default.
W2398093498 cites W2054892825 @default.
W2398093498 cites W2055828956 @default.
W2398093498 cites W2057221557 @default.
W2398093498 cites W2057369997 @default.
W2398093498 cites W2062429798 @default.
W2398093498 cites W2076617476 @default.
W2398093498 cites W2086995514 @default.
W2398093498 cites W2094031942 @default.
W2398093498 cites W2096716908 @default.
W2398093498 cites W2099977650 @default.
W2398093498 cites W2104909300 @default.
W2398093498 cites W2113578554 @default.
W2398093498 cites W2133444394 @default.
W2398093498 cites W2149003372 @default.
W2398093498 cites W2157003085 @default.
W2398093498 cites W2162268280 @default.
W2398093498 cites W2163201426 @default.
W2398093498 cites W2398665273 @default.
W2398093498 cites W4206154357 @default.
W2398093498 cites W4246362487 @default.
W2398093498 cites W2020277977 @default.
W2398093498 doi "https://doi.org/10.1542/neo.5-2-e33" @default.
W2398093498 hasPublicationYear "2004" @default.
W2398093498 type Work @default.
W2398093498 sameAs 2398093498 @default.
W2398093498 citedByCount "12" @default.
W2398093498 countsByYear W23980934982014 @default.
W2398093498 countsByYear W23980934982015 @default.
W2398093498 countsByYear W23980934982017 @default.
W2398093498 countsByYear W23980934982018 @default.
W2398093498 countsByYear W23980934982019 @default.
W2398093498 countsByYear W23980934982021 @default.
W2398093498 crossrefType "journal-article" @default.
W2398093498 hasAuthorship W2398093498A5007535386 @default.
W2398093498 hasConcept C111472728 @default.
W2398093498 hasConcept C138885662 @default.
W2398093498 hasConcept C177713679 @default.
W2398093498 hasConcept C187212893 @default.
W2398093498 hasConcept C2779234561 @default.
W2398093498 hasConcept C2779530757 @default.
W2398093498 hasConcept C2993955275 @default.
W2398093498 hasConcept C512399662 @default.
W2398093498 hasConcept C54355233 @default.
W2398093498 hasConcept C71924100 @default.
W2398093498 hasConcept C86803240 @default.
W2398093498 hasConceptScore W2398093498C111472728 @default.
W2398093498 hasConceptScore W2398093498C138885662 @default.
W2398093498 hasConceptScore W2398093498C177713679 @default.
W2398093498 hasConceptScore W2398093498C187212893 @default.
W2398093498 hasConceptScore W2398093498C2779234561 @default.
W2398093498 hasConceptScore W2398093498C2779530757 @default.
W2398093498 hasConceptScore W2398093498C2993955275 @default.
W2398093498 hasConceptScore W2398093498C512399662 @default.
W2398093498 hasConceptScore W2398093498C54355233 @default.
W2398093498 hasConceptScore W2398093498C71924100 @default.
W2398093498 hasConceptScore W2398093498C86803240 @default.
W2398093498 hasIssue "2" @default.
W2398093498 hasLocation W23980934981 @default.
W2398093498 hasOpenAccess W2398093498 @default.
W2398093498 hasPrimaryLocation W23980934981 @default.
W2398093498 hasRelatedWork W1506200166 @default.
W2398093498 hasRelatedWork W1995515455 @default.
W2398093498 hasRelatedWork W2048182022 @default.
W2398093498 hasRelatedWork W2080531066 @default.
W2398093498 hasRelatedWork W2463161041 @default.
W2398093498 hasRelatedWork W2748952813 @default.
W2398093498 hasRelatedWork W2899084033 @default.
W2398093498 hasRelatedWork W3031052312 @default.
W2398093498 hasRelatedWork W3032375762 @default.
W2398093498 hasRelatedWork W3108674512 @default.
W2398093498 hasVolume "5" @default.
W2398093498 isParatext "false" @default.
W2398093498 isRetracted "false" @default.
W2398093498 magId "2398093498" @default.
W2398093498 workType "article" @default.