Skip Navigation LinksHome > January 2014 - Volume 120 - Issue 1 > Preoperative Thrombocytopenia and Postoperative Outcomes aft...
Anesthesiology:
doi: 10.1097/ALN.0b013e3182a4441f
Perioperative Medicine

Preoperative Thrombocytopenia and Postoperative Outcomes after Noncardiac Surgery

Glance, Laurent G. M.D.; Blumberg, Neil M.D.; Eaton, Michael P. M.D.; Lustik, Stewart J. M.D.; Osler, Turner M. M.D.; Wissler, Richard M.D.; Zollo, Ray M.D.; Karcz, Marcin M.D.; Feng, Changyong Ph.D.; Dick, Andrew W. Ph.D.

Free Access
Article Outline
Collapse Box

Author Information

Collapse Box

Abstract

Background: Most studies examining the prognostic value of preoperative coagulation testing are too small to examine the predictive value of routine preoperative coagulation testing in patients having noncardiac surgery.
Methods: Using data from the American College of Surgeons National Surgical Quality Improvement database, the authors performed a retrospective observational study on 316,644 patients having noncardiac surgery who did not have clinical indications for preoperative coagulation testing. The authors used multivariable logistic regression analysis to explore the association between platelet count abnormalities and red cell transfusion, mortality, and major complications.
Results: Thrombocytopenia or thrombocytosis occurred in 1 in 14 patients without clinical indications for preoperative platelet testing. Patients with mild thrombocytopenia (101,000–150,000 µl−1), moderate-to-severe thrombocytopenia (<100,000 µl−1), and thrombocytosis (≥450,000 µl−1) were significantly more likely to be transfused (7.3%, 11.8%, 8.9%, 3.1%) and had significantly higher 30-day mortality rates (1.5%, 2.6%, 0.9%, 0.5%) compared with patients with a normal platelet count. In the multivariable analyses, mild thrombocytopenia (adjusted odds ratio [AOR], 1.28; 95% CI, 1.18–1.39) and moderate-to-severe thrombocytopenia (AOR, 1.76; 95% CI, 1.49–2.08), and thrombocytosis (AOR, 1.44; 95% CI, 1.30–1.60) were associated with increased risk of blood transfusion. Mild thrombocytopenia (AOR, 1.31; 95% CI, 1.11–1.56) and moderate-to-severe thrombocytopenia (AOR, 1.93; 95% CI, 1.43–2.61) were also associated with increased risk of 30-day mortality, whereas thrombocytosis was not (AOR, 0.94; 95% CI, 0.72–1.22).
Conclusion: Platelet count abnormalities found in the course of routine preoperative screening are associated with a higher risk of blood transfusion and death.
Back to Top | Article Outline

What We Already Know about This Topic

* The value of routine preoperative coagulation testing for noncardiac surgery remains unclear
* The investigators evaluated the association between platelet count abnormalities and transfusions and mortality in the National Surgical Quality Improvement Project registry
Back to Top | Article Outline

What This Article Tells Us That Is New

* Abnormal platelet counts were common, occurring in 1 of 14 patients
* Patients with low platelet counts were more likely to require transfusion and more likely to die
PATIENTS having surgery in the United States receive nearly one quarter of the 15 million units of blood transfused annually, at a cost of nearly 800 million dollars.1 Patients having noncardiac surgery who receive as few as 1–2 units of erythrocytes have a 30% higher risk of mortality (ROM) compared with patients who are not transfused.2,3 There is enormous variability in transfusion practices across hospitals for both cardiac and noncardiac surgery, with greater than four-fold differences in transfusion rates between hospitals for some procedures.4,5 Practice variation can be a red flag for low-quality health care and may be useful for identifying targets for quality improvement and cost reduction.6 In theory, coagulation testing may help preoperatively identify patients who are at increased risk for perioperative bleeding and who may benefit from targeted interventions to reduce bleeding and blood use.
In the past, coagulation testing was widely performed to identify patients at risk for bleeding complications during surgery. However, it is now widely accepted that routine preoperative testing is not clinically useful and is needlessly expensive.7 A recent work group convened by the American College of Physicians classified routine preoperative laboratory testing as inconsistent with “high-value, cost-conscious care.”8 According to the 2012 American Society of Anesthesiologists Task Force on Preanesthesia Evaluation, “preoperative tests should not be ordered routinely.”9 In particular, the Task Force recommends that the decision to order preoperative coagulation testing should be guided by the history and physical examination of the patient.9 This recommendation is supported by much evidence, accumulated over the past 30 yr, suggesting that routine coagulation testing is not clinically useful for identifying patients who are more likely to develop complications after surgery.10–19 However, most of these studies are based on relatively small patient cohorts, which may be inadequately powered to support such a conclusion.
The primary hypothesis of this exploratory study was that thrombocytopenic patients having noncardiac surgery are more likely to receive erythrocyte transfusions compared with patients with a normal platelet count. The secondary hypothesis was that thrombocytopenia is associated with a higher ROM and major complications. Our analysis only included patients without clinical indications for coagulation testing. The purpose of this study was to reexamine the predictive value of platelet testing in a large asymptomatic patient cohort in order to help inform future guideline development for coagulation testing.
Back to Top | Article Outline

Materials and Methods

Data Source
This study was conducted using data from the American College of Surgeons National Surgical Quality Improvement database (NSQIP) for patients having noncardiac surgery in 2008 and 2009. American College of Surgeons NSQIP was designed to promote quality improvement using performance feedback.20 NSQIP includes information on patient demographics, admission source, preoperative risk factors, intraoperative transfusion, and 30-day postoperative outcomes for patients having major surgery in more than 200 participating hospitals.21
The data do not include information on postoperative transfusions, except to indicate if a patient has received more than 4 units of erythrocytes postoperatively, nor do they include information on platelet transfusions. Data quality is ensured through the use of trained Surgical Clinical Reviewers and auditing of participating sites. Patient data are abstracted from the medical chart, operative log, anesthesia record, from interviews with the attending surgeon, and telephone interviews with patients. A systematic sampling strategy is used to avoid bias in case selection, and to ensure a diverse surgical case mix.21 The University of Rochester School of Medicine Institutional Review Board (Rochester, New York) determined that this study did not require Institutional Review Board review.
Back to Top | Article Outline
Study Population and Outcomes
We identified 514,087 patients who underwent general, vascular, or orthopedic surgery under general, spinal, or epidural anesthesia using Current Procedural Terminology codes.
Fig. 1
Fig. 1
Image Tools
From this sample, we identified patients without any one of the following clinical indications for preoperative platelet testing12: self-reported history of a bleeding disorder (vitamin K deficiency, hemophilia, thrombocytopenia, chronic anticoagulation therapy that has not been discontinued before surgery), hemorrhage (preoperative transfusion of >4 units of blood), chemotherapy in the 30 days before surgery, radiation therapy within 90 days of surgery, severe infections (systemic inflammatory response syndrome, sepsis, septic shock), liver disease (ascites, esophageal varices), steroid use, malnutrition (>10% decrease in body weight in the 6 months before surgery), renal failure, or disseminated cancer (405,343). We excluded records with no platelet counts (71,276), no hematocrits (838), procedures with work relative value units equal to zero (7,790), missing demographic information (7,780), missing American Society of Anesthesiologist Physical Status (428), and missing information on blood transfusion (547), to yield an analytic sample of 316,644 records of patients without clinical indications for coagulation testing (fig. 1). Eighty-five percent of the patients with missing platelet counts were also missing a hematocrit. It was impossible to determine whether patient records without platelet count or hematocrit information were due to patient testing not being done or due to missing data.
Back to Top | Article Outline
Statistical Analysis
The primary outcome was the receipt of any erythrocyte transfusion. Secondary outcomes were 30-day mortality and the following 30-day complications: (1) cardiac (acute myocardial infarction or cardiac arrest); (2) pulmonary (pneumonia, ventilatory support for >48 h); (3) renal (progressive renal insufficiency or acute renal failure); (4) central nervous system (cerebrovascular accident or coma lasting >24 h); (5) sepsis (sepsis or septic shock); (6) wound infection (deep incisional surgical site infection, organ or space surgical site infection, or wound dehiscence*); (6) thromboembolic (deep venous thrombosis or pulmonary embolism); and (7) graft failure.
Categories of Proced...
Categories of Proced...
Image Tools
A priori we divided the preoperative platelet count into the following categories: (1) moderate-to-severe thrombocytopenia (<100,000 µl−1); (2) mild thrombocytopenia (101,000–150,000 µl−1); (3) low-normal (151,000–200,000 µl−1); (4) normal (201,000–450,000 µl−1); and (5) thrombocytosis (≥450,000 µl−1). We first performed exploratory analyses to compare patients with mild thrombocytopenia, moderate-to-severe thrombocytopenia, low-normal platelet count, normal platelet count, and increased platelet count using chi-square analysis, linear regression, or logistic regression, as appropriate. We then estimated the independent association between receipt of any erythrocyte transfusion and platelet count, using multivariable logistic regression, controlling for hematocrit, age, sex, body mass index (underweight, overweight, obesity, morbid obesity, and super obesity), admission source (home, transfer from other hospital, chronic care facility), race, inpatient status (vs. outpatient), emergency status, surgical complexity (work relative value units), previous operation within 30 days, and comorbidities: diabetes (oral hypoglycemics, insulin treatment), pulmonary (chronic obstructive pulmonary disease, pneumonia, mechanical ventilation before surgery, dyspnea at rest, dyspnea on exertion), cardiac (congestive heart failure, myocardial infarction, angina, percutaneous coronary intervention, open heart surgery), hypertension, peripheral vascular disease, renal disease (stage 2 chronic kidney disease: glomerular filtration rate, 60–89 ml·min·1.73 m−2; stage 3 chronic kidney disease: glomerular filtration rate, 30–59 ml·min·1.73 m−2; stage 4 chronic kidney disease: glomerular filtration rate, 15–29 ml·min·1.73 m−2),22 central nervous system (stroke with neurologic deficit, stroke without neurologic deficit, transient ischemic attack, impaired sensorium, coma, hemiplegia, paraplegia, quadriplegia, tumor involving the central nervous system). In addition to work relative value units as a measure of surgical complexity, we included separate intercept terms for the type of procedure by Current Procedural Terminology code group (appendix 1): (1) integumentary; (2) musculoskeletal; (3) vascular; (4) hemic and lymphatic system; (5) mouth, palate, salivary glands, pharynx, adenoids, and esophagus; (6) stomach, intestines, appendix and mesentery, rectum and anus, liver, biliary tract, pancreas, abdomen, peritoneum, and omentum (nonhernia); (8) endocrine system; and (9) hernia repair (reference group).
Stratified analyses were performed in (1) low-risk patients (defined as having an ROM ≤0.5%; intermediate-risk patients (ROM: >0.5 to 3.5%); and high-risk (ROM: >3.5%). ROM was estimated using a logistic regression model based on the set of clinical risk factors described above (see Statistical Analysis, paragraph 2), omitting platelet count.
In our secondary analyses, we estimated the independent association between platelet count and (1) mortality; (2) mortality or any major complication; and (3) individual complications. In the baseline models, we used the same list of risk factors listed above. We then reestimated each model, also including blood transfusion as an additional covariate, to determine whether the effect size of the platelet count on patient outcomes would be reduced when accounting for the effects of blood transfusion. Blood transfusion was specified as a categorical variable: 0 unit erythrocytes – (reference category, 1 unit erythrocytes, 2 units erythrocytes, 3 units erythrocytes, 4 units erythrocytes, and >4 units erythrocytes).
We did not perform a sensitivity analysis to examine the association between missing platelet count and blood transfusion because 85% of patients with missing platelet counts (n = 71,276) were also missing hematocrit values. Presumably, if clinicians elect to not order a platelet count in patients without clinical indications for platelet testing, then those patients with a missing platelet count would be less likely to require blood transfusion. Performing a sensitivity analysis in which we include patients with missing platelet counts as a separate category would not be very informative because most of these patients are also missing one of the most important confounders, the hematocrit. Without information on baseline hematocrit, this sensitivity analysis would be potentially biased. The alternative, which is to include patients with missing platelet counts and impute missing values for the platelet count (and for the hematocrit) would not be appropriate, because platelet count is the exposure variable in our analysis.
Data management and statistical analyses were performed using STATA SE/MP version 12 (STATA Corp., College Station, TX). We used multiple imputation to impute missing values for preoperative serum creatinine using the STATA implementation of the multiple imputation by chained equations method of multiple imputation23 described by van Buuren et al.24 We specified the imputation model using nonparsimonious linear regression estimated in the entire analytic sample. Rubin rule was used to combine parameter estimates across the five imputed data sets obtained by multiple imputation.23 Model discrimination was assessed using the C statistic. All statistical tests were two-tailed and P values less than 0.05 were considered significant. We used robust variance estimators to account for the nonindependence of observations within hospitals.25
Back to Top | Article Outline

Results

Descriptive Statistics
Table 1
Table 1
Image Tools
The analytic sample consisted of 316,644 patients. Patients with severe thrombocytopenia tended to be older (62 vs. 55), female (56 vs. 35%), have lower hematocrits (37.2 vs. 39.9), and transferred from another hospital (2.62 vs. 0.86%) compared with patients with a normal platelet count (table 1). In general, severely thrombocytopenic patients were more likely to have a history of previous cardiac surgery (11.4 vs. 3.83%), percutaneous coronary intervention (7.14 vs. 3.91%), moderate (24.5 vs. 14.3%) or severe renal disease (3.84 vs. 1.39%), and chronic obstructive pulmonary disease (5.70 vs. 3.72%). Patients with severe thrombocytopenia were more likely to be at high risk (8.71 vs. 2.32%) and intermediate risk (39.2 vs. 24.0%) for 30-day mortality compared with patients with normal platelet counts.
Back to Top | Article Outline
Unadjusted Analyses
Table 2
Table 2
Image Tools
Fig. 2
Fig. 2
Image Tools
Unadjusted Outcomes,...
Unadjusted Outcomes,...
Image Tools
In the unadjusted analyses, moderate-to-severe thrombocytopenia (platelet count ≤100,000) was associated with a four-fold higher odds of receiving any transfusion; transfusion rate, 11.8%; odds ratio [OR], 4.15; 95% CI, 3.66–4.71; P ≤ 0.001), and mild thrombocytopenia (platelet count, 101,000–150,000) with a 2.5-fold higher risk of receiving any transfusion (observed transfusion rate, 7.3%; OR, 2.45; 95% CI, 2.28–2.63; P ≤ 0.001), compared with patients with normal platelet counts (observed transfusion rate, 3.1%; fig. 2, table 2, and appendix 2). Patients with thrombocytosis (platelet count, ≥450,000) had a three-fold higher unadjusted odds of receiving any blood transfusion (transfusion rate, 8.9%; OR, 3.02; 95% CI, 2.78–3.28; P ≤ 0.001) compared with patients with a normal platelet count.
Table 3
Table 3
Image Tools
Fig. 3
Fig. 3
Image Tools
In the unadjusted analyses, patients with either mild (observed mortality rate 1.5%; OR, 3.18; 95% CI, 2.73–3.70; P ≤ 0.001) or moderate-to-severe thrombocytopenia (observed mortality rate, 2.6%; OR, 5.49; 95% CI, 4.24–7.19; P ≤ 0.001) were more likely to die compared with patients with normal platelet counts (observed mortality rate, 0.5%; fig. 3, table 3, and appendix 2). Patients with either mild (mortality or complication rate, 8.4%; OR, 1.71; 95% CI, 1.60–1.83; P ≤ 0.001) or moderate-to-severe thrombocytopenia (mortality or complication rate, 12.2%; OR, 2.61; 95% CI, 2.31–2.96; P ≤ 0.001) were also more likely to die or have a major complication compared with patients with normal platelet counts (mortality or complication rate, 5.1%). Patients with thrombocytosis were also at higher risk for death (observed mortality rate, 0.9%; OR, 1.81; 95% CI, 1.41–2.32; P ≤ 0.001), or death or major complications (mortality or complication rate, 11.0%; OR, 2.32; 95% CI, 2.15–2.50; P ≤ 0.001).
Patients with missing values for platelet count and/or hematocrit had lower rates of blood transfusion (0.9%), death (0.1%), and death or major complications (1.7%) compared with patients in the analytic sample (appendix 2). Patients with missing values for the platelet count and/or hematocrit were also more likely to have American Society of Anesthesiologist Physical Status I (18.2 vs. 8.9%) and American Society of Anesthesiologist Physical Status II (59.2 vs. 51.0%) compared with patients with normal platelet counts.
Back to Top | Article Outline
Results of Multivariable Analyses
Model Parameters for...
Model Parameters for...
Image Tools
We present the results of our analyses based on multiple imputation of missing values for creatinine (missing in 5.4% of the analytic sample). In the multivariable analysis, patients with mild (adjusted OR [AOR], 1.29; 95% CI, 1.18–1.39; P ≤ 0.001) or moderate-to-severe thrombocytopenia (AOR, 1.76; 95% CI, 1.49–2.08; P ≤ 0.001) were more likely to receive a blood transfusion compared with patients with normal platelet counts (table 2 and appendix 3). Patients with thrombocytosis (AOR, 1.44; 95% CI, 1.30–1.60; P ≤ 0.001) were also more likely to receive blood compared with patients with normal platelet counts. Similar results were obtained when these analyses were repeated in separate strata of low-risk, intermediate-risk, and high-risk patients (table 2 and appendix 3).
Thrombocytopenia and thrombocytosis are also associated with a higher ROM. Mild thrombocytopenia (AOR, 1.31; 95% CI, 1.11–1.56; P ≤ 0.01) and moderate-to-severe thrombocytopenia (AOR, 1.93; 95% CI, 1.43–2.61; P ≤ 0.001) were significantly associated with a higher ROM, whereas thrombocytosis was not (AOR, 0.94; 95% CI, 0.72–1.23; P = 0.60; table 3 and appendix 3). We also found that platelet count abnormalities were associated with a higher risk of the composite outcome—death or major complications. Patients with moderate-to-severe thrombocytopenia or thrombocytosis were more likely to have a major complication or die: moderate-to-severe thrombocytopenia (AOR, 1.52; 95% CI, 1.32–1.74; P ≤ 0.001); mild thrombocytopenia (AOR, 1.12; 95% CI, 1.04–1.20; P ≤ 0.01); or thrombocytosis (AOR, 1.36; 95% CI, 1.25–1.48; P ≤ 0.001; table 3).
The inclusion of intraoperative and postoperative transfusions reduced the association between platelet count abnormalities and death, or death and major complications (table 3). Results obtained in stratified analyses were qualitatively similar to those obtained in the entire sample (table 3).
Table 4
Table 4
Image Tools
We also examined the association between individual complications and outcomes. Platelet count abnormalities were associated with higher risk of sepsis, pulmonary, renal, wound, and thromboembolic complications (table 4). Patients with moderate-to-severe thrombocytopenia were more likely to experience sepsis (AOR, 1.27; 95% CI, 1.12–1.45; P ≤ 0.001), adverse pulmonary outcomes (AOR, 1.87; 95% CI, 1.50–2.32; P ≤ 0.001), and renal complications (AOR, 2.05; 95% CI, 1.48–2.85; P ≤ 0.001) whereas patients with thrombocytosis were more likely to experience pulmonary (AOR, 1.30; 95% CI, 1.12–1.52; P ≤ 0.001), renal (AOR, 1.48; 95% CI, 1.14–1.91; P ≤ 0.01), wound (AOR, 1.49; 95% CI, 1.31–1.69; P ≤ 0.001), and thromboembolic complications (AOR, 1.74; 95% CI, 1.43–2.11; P ≤0.001; table 4). All models exhibited acceptable discrimination.
Back to Top | Article Outline

Discussion

In this large observational study of more than 300,000 noncardiac surgical patients without known indications for preoperative coagulation testing, we found that patients with moderate-to-severe thrombocytopenia had a 75% higher risk of receiving a blood transfusion, and a 90% higher risk of 30-day mortality, after adjusting for preoperative risk and surgical complexity. Patients with mild thrombocytopenia had a 29% increased risk of transfusion and 31% higher ROM. Thrombocytosis was associated with a 44% increased risk of transfusion, but no increase in mortality. Thrombocytopenia and thrombocytosis were also associated with an increased risk of pulmonary, renal, sepsis, wound, and thromboembolic complications. Thrombocytopenia or thrombocytosis were not uncommon, occurring in 1 in 14 patients without clinical indications for preoperative platelet testing. To our knowledge, this is the largest study to date to examine the association between platelet count and perioperative outcomes in noncardiac surgical patients without clinical indications for coagulation testing.
The ability of previous studies to evaluate the predictive value of preoperative coagulation testing was limited by low sample sizes. In a study based on 92 patients having cardiac operations, Ramsey et al.11 did not find a difference in transfusion requirements in patients with abnormal preoperative coagulation tests. Rohrer et al.14 prospectively studied 163 general and vascular surgery patients who received screening coagulation tests, 13 of whom had abnormal platelet counts, none of which were associated with adverse events. In a single-center retrospective study of 2,000 elective surgical patients, 60% of preoperative testing was designated as routine screening; only four of the test results were considered abnormal and none influenced perioperative care.12 In a retrospective case-control study of 828 patients having noncardiac surgery, Ng et al.18 found that coagulation abnormalities were associated with blood transfusion, but that nearly all coagulation abnormalities occurred in patients with clinical indications for testing. Houry et al.16 conducted a prospective multicenter study of 3,242 general surgery patients, in which they found no changes in blood use, blood drainage, reoperations to control hemorrhage, and mortality in the 340 patients with abnormal coagulation tests and no clinical abnormalities, compared with patients with normal coagulation parameters. In a prospective cohort of 544 elderly noncardiac surgical patients, abnormalities in platelet count occurred in 10 patients and were not associated with postoperative adverse outcomes.17 In a single-center pilot study designed to evaluate whether indicated preoperative testing can be eliminated in ambulatory surgery, 1,026 patients were randomized to indicated preoperative testing or no testing. This pilot study did not find significant differences in adverse events between the testing and no-testing group.19 In comparison, a larger, more recent study of 11,804 adult neurosurgery patients using the American College of Surgeons NSQIP database found that coagulation abnormalities were significantly associated with blood transfusion, reoperation, and mortality.26 This study did not, however, examine the prognostic value of coagulation testing in patients without clinical indications for coagulation testing, and only included neurosurgical patients. With the exception of the last study, most of these studies appear to lack the statistical power to clearly establish the prognostic importance of platelet testing in surgical patients.
Our study cannot establish a causal relationship between platelet count abnormalities and death and major complications in patients having noncardiac surgery. However, our findings have face validity. Thrombocytopenic patients are more likely to bleed after surgery,27 and patients who bleed excessively have worse outcomes. In our study, thrombocytopenic patients were 30–80% more likely to receive blood transfusion than those with normal platelet counts. Accounting for the effects of blood transfusion led to smaller increases in mortality and complications in thrombocytopenic patients. However, we could not accurately estimate the contribution of blood transfusion to the adverse effects of thrombocytopenia because we lacked detailed information on postoperative blood transfusion. There are many causes of thrombocytopenia, including immune-mediated diseases, bone marrow disorders, and platelet sequestration.27 In some cases, platelet abnormalities may lead to adverse outcomes after surgery, independent of the role of platelets in hemostasis, and may instead serve as a marker for other systemic diseases that impact postoperative outcomes. Our findings add to the accumulating evidence that hematologic abnormalities are important risk factors in surgical patients.28,29
This study has several important limitations. Although we designed this study to examine the prognostic importance of routine platelet count testing, and thus purposefully excluded patients with clinical indications for coagulation testing from our analysis, we cannot completely rule out the possibility that some of the patients who underwent coagulation testing in our sample may have had clinical indications for testing, which were not captured in our database. In addition, our analysis could only examine the prognostic value of the platelet count in patients who actually had coagulation testing, and not in the sample of patients (18% of our original study sample) who were excluded from our analysis because they did not have a recorded platelet count. These patients were less likely to be transfused or to have poor outcomes, and were also more likely to be American Society of Anesthesiologist Physical Status I or II. It is thus possible that our findings overestimate the prognostic importance of platelet testing. This concern is mitigated by the finding that although patients excluded from our analysis due to missing platelet counts were less likely to be transfused, they were also generally healthier compared with patients with normal platelet counts.
The retrospective nature of this study allows us only to conclude that there is an association between platelet count abnormalities and adverse events. We cannot conclude that platelet count abnormalities resulted in worse outcomes. More importantly, our study does not allow us to conclude that in the absence of routine coagulation screening, clinicians would have delivered less-effective care resulting in more adverse outcomes. Like other observational studies, this study may be biased due to unmeasured confounding, despite the comprehensiveness of the American College of Surgeons NSQIP database. Missing values for the serum creatinine were imputed using multiple imputation. This approach is less likely to cause bias than complete case analysis in which observations with missing data are excluded.30 Because the public use NSQIP data file does not include hospital identifiers, we could not control for hospital effects. However, differences in hospital quality would only bias our analysis if patients with abnormal platelet counts were more likely to be admitted to lower-quality hospitals. Finally, because our data do not include information on platelet transfusions, we were unable to examine the association between platelet testing and platelet transfusion.
Given the imperative to cut healthcare costs, major stakeholders are launching initiatives to reduce unnecessary testing.8 Our findings lead us to question whether recent recommendations9 discouraging preoperative testing are too restrictive and perhaps based on insufficient evidence. Likewise, our study results must be interpreted with appropriate reservation, given the retrospective nature of our study. Nonetheless, healthcare systems are inherently complex and prone to patient safety errors. Although in theory, thorough histories and well-conducted physical examinations may identify many patients who will benefit the most from selective preoperative testing, a less-selective approach to coagulation testing adds some redundancy to preoperative evaluations and may identify high-risk patients who might be otherwise “missed.” Given the production pressures of medical practice and the common delegation of preoperative screening to nonphysicians,31 finding the appropriate balance between routinely testing everyone and selectively testing will require careful study.
If less-restrictive coagulation testing identifies a group of high-risk patients who would be missed if more traditional screening criteria were used, can this information be used to improve patient outcomes? It is easier to identify risk factors for poor outcomes than it is to identify effective risk-reduction strategies.32 However, one can speculate that preoperative screening tests can improve shared decision making. Although the decision to undergo surgery is frequently driven by patient preferences, patients often lack the necessary knowledge about the risks and benefits of surgery to engage in shared decision making.33 Truly patient-centered medicine requires that we inform patients of their personal risk of surgery based on their own personal set of risk factors, as opposed to a more general discussion based on the “typical” patient. Providing patients with personal risk information may help advance patient-centered medicine as “the next phase in health care,”32 by allowing patients and their physicians to more accurately weigh the risks and benefits of surgery.
Back to Top | Article Outline

Conclusion

Platelet count abnormalities found in the course of “routine” preoperative screening are associated with a higher risk of blood transfusion, major complications, and death. We thus question whether recent recommendations9 discouraging preoperative testing are too restrictive and perhaps based on insufficient evidence.
* Superficial surgical site infections were not included. Cited Here...
Back to Top | Article Outline

References

1. The 2009 National Blood Collection and Utilization Survey. 2009 Washington, DC: Department of Health and Human Services

2. Bernard AC, Davenport DL, Chang PK, Vaughan TB, Zwischenberger JB. Intraoperative transfusion of 1 U to 2 U packed erythrocytes is associated with increased 30 day mortality, surgical-site infection, pneumonia, and sepsis in general surgery patients. J Am Coll Surg. 2009;208:931–7

3. Glance LG, Dick AW, Mukamel DB, Fleming FJ, Zollo RA, Wissler R, Salloum R, Meredith UW, Osler TM. Association between intraoperative blood transfusion and mortality and morbidity in patients undergoing noncardiac surgery. A. 2011;114:283–92

4. Bennett-Guerrero E, Zhao Y, O’Brien SM, Ferguson TB Jr, Peterson ED, Gammie JS, Song HK. Variation in use of blood transfusion in coronary artery bypass graft surgery. JAMA. 2010;304:1568–75

5. Qian F, Osler TM, Eaton MP, Dick AW, Hohmann SF, Lustik SJ, Diachun CA, Pasternak R, Wissler RN, Glance LG. Variation of blood transfusion in patients undergoing major noncardiac surgery. Ann Surg. 2013;257:266–78

6. Bernheim SM, Grady JN, Lin Z, Wang Y, Wang Y, Savage SV, Bhat KR, Ross JS, Desai MM, Merrill AR, Han LF, Rapp MT, Drye EE, Normand SL, Krumholz HM. National patterns of risk-standardized mortality and readmission for acute myocardial infarction and heart failure. Update on publicly reported outcomes measures based on the 2010 release. Circ Cardiovasc Qual Outcomes. 2010;3:459–67

7. García-Miguel FJ, Serrano-Aguilar PG, López-Bastida J. Preoperative assessment. Lancet. 2003;362:1749–57

8. Qaseem A, Alguire P, Dallas P, Feinberg LE, Fitzgerald FT, Horwitch C, Humphrey L, LeBlond R, Moyer D, Wiese JG, Weinberger S. Appropriate use of screening and diagnostic tests to foster high-value, cost-conscious care. Ann Intern Med. 2012;156:147–9

9. Apfelbaum JL, Connis RT, Nickinovich DG, Pasternak LR, Arens JF, Caplan RA, Fleisher LA, Flowerdew R, Gold BS, Mayhew JF, Rice LJ, Roizen MF, Twersky RS. Practice advisory for preanesthesia evaluation: An updated report by the American Society of Anesthesiologists Task Force on Preanesthesia Evaluation. A. 2012;116:522–38

10. Eisenberg JM, Clarke JR, Sussman SA. Prothrombin and partial thromboplastin times as preoperative screening tests. Arch Surg. 1982;117:48–51

11. Ramsey G, Arvan DA, Stewart S, Blumberg N. Do preoperative laboratory tests predict blood transfusion needs in cardiac operations? J Thorac Cardiovasc Surg. 1983;85:564–9

12. Kaplan EB, Sheiner LB, Boeckmann AJ, Roizen MF, Beal SL, Cohen SN, Nicoll CD. The usefulness of preoperative laboratory screening. JAMA. 1985;253:3576–81

13. Suchman AL, Mushlin AI. How well does the activated partial thromboplastin time predict postoperative hemorrhage? JAMA. 1986;256:750–3

14. Rohrer MJ, Michelotti MC, Nahrwold DL. A prospective evaluation of the efficacy of preoperative coagulation testing. Ann Surg. 1988;208:554–7

15. Erban SB, Kinman JL, Schwartz JS. Routine use of the prothrombin and partial thromboplastin times. JAMA. 1989;262:2428–32

16. Houry S, Georgeac C, Hay JM, Fingerhut A, Boudet MJ. A prospective multicenter evaluation of preoperative hemostatic screening tests. The French Associations for Surgical Research. Am J Surg. 1995;170:19–23

17. Dzankic S, Pastor D, Gonzalez C, Leung JM. The prevalence and predictive value of abnormal preoperative laboratory tests in elderly surgical patients. Anesth Analg. 2001;93:301–8

18. Ng KF, Lai KW, Tsang SF. Value of preoperative coagulation tests: Reappraisal of major noncardiac surgery. World J Surg. 2002;26:515–20

19. Chung F, Yuan H, Yin L, Vairavanathan S, Wong DT. Elimination of preoperative testing in ambulatory surgery. Anesth Analg. 2009;108:467–75

20. Hall BL, Hamilton BH, Richards K, Bilimoria KY, Cohen ME, Ko CY. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: An evaluation of all participating hospitals. Ann Surg. 2009;250:363–76

21. Khuri SF, Henderson WG, Daley J, Jonasson O, Jones RS, Campbell DA Jr, Fink AS, Mentzer RM Jr, Steeger JEPrincipal Site Investigators of the Patient Safety in Surgery Study. . The patient safety in surgery study: Background, study design, and patient populations. J Am Coll Surg. 2007;204:1089–102

22. Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, Hogg RJ, Perrone RD, Lau J, Eknoyan GNational Kidney Foundation. . National Kidney Foundation practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Ann Intern Med. 2003;139:137–47

23. Multiple Imputation of Missing Values. Stata J 2004; 4:227-41

24. van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing blood pressure covariates in survival analysis. Stat Med. 1999;18:681–94

25. White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica. 1980;48:817–30

26. Seicean A, Schiltz NK, Seicean S, Alan N, Neuhauser D, Weil RJ. Use and utility of preoperative hemostatic screening and patient history in adult neurosurgical patients. J Neurosurg. 2012;116:1097–105

27. Abrams CSGoldman L. Thrombocytopenia Goldman’s Cecil Medicine. 2011 Philadelphia Saunders:1124–31 . Edited by

28. Wu WC, Schifftner TL, Henderson WG, Eaton CB, Poses RM, Uttley G, Sharma SC, Vezeridis M, Khuri SF, Friedmann PD. Preoperative hematocrit levels and postoperative outcomes in older patients undergoing noncardiac surgery. JAMA. 2007;297:2481–8

29. Musallam KM, Tamim HM, Richards T, Spahn DR, Rosendaal FR, Habbal A, Khreiss M, Dahdaleh FS, Khavandi K, Sfeir PM, Soweid A, Hoballah JJ, Taher AT, Jamali FR. Preoperative anaemia and postoperative outcomes in non-cardiac surgery: A retrospective cohort study. Lancet. 2011;378:1396–407

30. van der Heijden GJ, Donders AR, Stijnen T, Moons KG. Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: A clinical example. J Clin Epidemiol. 2006;59:1102–9

31. Yen C, Tsai M, Macario A. Preoperative evaluation clinics. Curr Opin Anaesthesiol. 2010;23:167–72

32. Krumholz HM. Patient-centered medicine: The next phase in health care. Circ Cardiovasc Qual Outcomes. 2011;4:374–5

33. Lee CN, Ko CY. Beyond outcomes—The appropriateness of surgical care. JAMA. 2009;302:1580–1

Back to Top | Article Outline
Appendix 1.
Back to Top | Article Outline
Appendix 2.
Back to Top | Article Outline
Appendix 3.

© 2014 American Society of Anesthesiologists, Inc.

Publication of an advertisement in Anesthesiology Online does not constitute endorsement by the American Society of Anesthesiologists, Inc. or Lippincott Williams & Wilkins, Inc. of the product or service being advertised.
Login

Article Tools

Images

Share