- Email: [email protected]
Abnormally High, as Well as Low, Preoperative Platelet Counts Correlate With Adverse Outcomes and Readmissions After Elective Total Knee Arthroplasty
The Journal of Arthroplasty 34 (2019) 1670e1676
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Primary Arthroplasty
Abnormally High, as Well as Low, Preoperative Platelet Counts Correlate With Adverse Outcomes and Readmissions After Elective Total Knee Arthroplasty Rohil Malpani a, Monique S. Haynes, MPH a, Michael G. Clark b, Anoop R. Galivanche a, Patawut Bovonratwet a, Jonathan N. Grauer, MD a, * a b
Department of Orthopaedics and Rehabilitation, Yale School of Medicine, New Haven, Connecticut Vanderbilt University School of Medicine, Nashville, Tennessee
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 December 2018 Received in revised form 26 March 2019 Accepted 6 April 2019 Available online 18 April 2019
Background: Laboratory studies are routinely performed as a part of the preoperative workup for a total knee arthroplasty (TKA). The ramifications of abnormal preoperative platelet counts remain uncharacterized in large, multicenter patient populations. Methods: Patients who underwent elective primary TKA were identified in the 2011-2015 National Surgical Quality Improvement Program database. Risk of 30-day postoperative complications was calculated as a function of preoperative platelet counts. Patients were characterized as having a normal platelet count, abnormally low platelet count, and abnormally high platelet count based on relative risk calculations. Univariate and multivariate analyses were performed to associate abnormal platelet counts with patient demographics, operative variables, 30-day postoperative complications, and readmissions. Results: In total, 140,073 patients who underwent elective TKA were identified. Using the relative risk threshold of 1.5 for any adverse event, abnormally low and abnormally high platelet count thresholds were set at 116,000/mL and 492,000/mL, respectively. Multivariate analyses revealed low platelet counts to be associated with higher rates of any, major, and minor adverse events and longer length of stay. Analogously, high platelet counts were associated with higher rates of any and minor adverse events and longer length of stay. Conclusion: The present study employed a large patient sample size and showed that elective TKA patients with abnormally high, as well as low, platelet counts are at increased risk of postoperative adverse outcomes. Focused attention needs to be paid to TKA patients with preoperative abnormal platelet counts for optimization and postoperative care. Level of Evidence: Level III, retrospective comparative study. © 2019 Elsevier Inc. All rights reserved.
Keywords: total knee arthroplasty the American College of Surgeons National Surgical Quality Improvement Program platelet count surgical outcomes postoperative adverse events readmission
Total knee arthroplasty (TKA) is a common and reliable surgical procedure for patients with degenerative knee conditions and functional impairment refractory to conservative treatment [1,2]. Disclaimer: Financial remuneration to authors and family members related to the subject of this article: None. One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to https://doi.org/10.1016/j.arth.2019.04.012. * Reprint requests: Jonathan N. Grauer, MD, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, 47 College Street, New Haven, CT 06520. https://doi.org/10.1016/j.arth.2019.04.012 0883-5403/© 2019 Elsevier Inc. All rights reserved.
More than 600,000 TKA procedures are performed annually in the United States, and regression models predict demand for primary TKA will exceed 3 million procedures annually by 2030 [3,4]. Optimizing outcomes after TKA is critical given the increase in surgical volume. Rising demand for TKA could place a financial burden on patients and hospital systems [3]. In the United States, the Bundled Payments for Care Improvement and Comprehensive Care for Joint Replacement models have been proposed to increase quality and coordination of care while reducing costs to Medicare [5,6]. Under these models, hospitals have a financial incentive to increase quality of care (ie, improve outcomes) and minimize cost (ie, reduce the frequency of surgical complications) [7]. These motivations
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
heighten the need for preoperative risk stratification models which could be used by providers to identify higher-risk surgeries and take steps to mitigate risk. Several qualitative and quantitative metrics for preoperative risk stratification have been evaluated. Patient age, for example, has been associated with increased mortality [8] and incidence of postoperative complications [8,9]. Similarly, high body mass index (BMI) has been associated with increased incidence of perioperative and postoperative complications [8,10], increased postoperative infection risk [11e15], and increased rates of revision surgery [12]. The presence of preexisting comorbidities, such as diabetes and cardiac disease, has been associated with increased risk of mortality [8], surgical complications [8,9], and infection [13,15]. Overall medical status also appears to be associated with surgical outcomes: American Society of Anesthesiologists (ASA) classifications 3 have been associated with higher incidence of postoperative complications [8] and infection [13]. Preoperative laboratory studies, such as chemistry panels, complete blood counts, and coagulation studies, are prime candidates for use in risk stratification models due to their ubiquity in the standard presurgical workup. Previous work has shown that low serum albumin levels are a strong predictor of surgical mortality and morbidity (especially sepsis and major infections) [16]. In cardiac surgery, increased platelet-to-lymphocyte ratio was shown to be associated with mortality and morbidity [17]. In noncardiac surgery, abnormal preoperative platelet counts have been associated with a higher incidence of blood transfusion and higher risk of death [18]. Other analyses have not found utility in certain preoperative laboratory studies. Serum D-dimer, for example, was shown to be an ineffective screening test for postoperative deep vein thrombosis following TKA [19]. Abnormal electrolyte, hemoglobin, creatinine, and glucose values, as well as thrombocytopenia, were not predictive of postoperative adverse outcomes in a study of elderly patients undergoing noncardiac surgery [20]. Other studies have suggested that routine preoperative laboratory screening tests contribute little to patient care and decision-making [21e23] and that not obtaining them does not adversely affect patient safety [24]. Many of the aforementioned positive and negative studies suffer from low cohort size (<1000) [17,19,20] or are not specific to TKA [16e18,20e24]. Consequently, the usefulness and clinical implications of abnormal preoperative platelet counts in patients undergoing primary TKA have not been fully established. As a laboratory test frequently ordered preoperatively, there is a need to aid in the interpretation of platelet count values before surgery, as well as establish relationships between preoperative values and postoperative events. To address this, the present study queried the National Surgical Quality Improvement Program (NSQIP) database to identify >100,000 patients who underwent TKA between 2011 and 2015. These data were interrogated to quantitatively characterize the range of platelet counts for patients undergoing primary TKA and to investigate associations between abnormal preoperative platelet counts and postoperative complications.
1671
following the procedure, regardless of discharge status. This database has been used extensively in various surgical specialty research, and its use in the orthopedic field has been extensive, dramatically increasing in recent years with higher-volume data [26e30]. Our institutional review board granted an exemption for studies using this dataset. Inclusion/Exclusion Criteria Patients who underwent elective primary TKA between 2011 and 2015 were identified using the Current Procedural Terminology code 27447. Procedures before 2011 did not include 30-day readmission data and were therefore not considered for the present study. Exclusion criteria included procedures performed for fracture, neoplasm, and infection. Patients who had a revision TKA (Current Procedural Terminology code 27486/27487) or who underwent nonelective TKA were also excluded. After exclusion, based on these criteria, a small proportion (<1%) of patients had missing data for vital demographic attributes and were excluded. An additional 5.25% of the patients had missing preoperative platelet count values and were also excluded. Of note, it has been shown in the orthopedic literature that variables with <10% missing data do not appreciably skew results of complete case analysis [31].
Materials and Methods
Patient Preoperative Characteristics and Perioperative Complications Age, sex, height, weight, functional status before surgery, ASA classification, smoking status, diabetes mellitus status, and preoperative platelet count were queried and directly abstracted from the dataset. Using height and weight data, BMI (in kg/m2) was calculated. ASA was used as a marker of comorbidity. Operating time and hospital length of stay (LOS) were abstracted from the database. Operating time (incision to closure) was measured in minutes while LOS (admission to discharge) was measured in days. The maximum value of the LOS was limited to 30 days. The NSQIP database contains data describing adverse outcomes from the 30-day follow-up period. Individual adverse events and aggregated any, major, and minor adverse events were assessed. The occurrence of a major adverse event was defined as the occurrence of at least one of the following: death, return to operating room, sepsis/septic shock, unplanned intubation, ventilator use >48 hours, stroke/cerebrovascular accident, cardiac arrest, acute renal failure, thromboembolic event (defined as pulmonary embolism or deep vein thrombosis), wound infection, return to the operating room, and 30-day readmissions. The occurrence of a minor adverse event was defined as the occurrence of any of the following: wound dehiscence, urinary tract infection, pneumonia, progressive renal insufficiency, and transfusions (a transfusion is defined in NSQIP as occurrence of transfusion of at least 1 unit of packed red blood cells starting from time of onset of the procedure up to 72 hours postoperatively, including hanging blood from the operating room that is finished outside the operating room, except NSQIP data from 2011 that did not collect intraoperative blood transfusion data). Any adverse event was defined as the occurrence of either a major or minor adverse event.
Database and Patient Population
Statistical Analysis
Database The NSQIP database contains over 300 variables for each surgical case from more than 500 participating healthcare institutions, the majority of which are based in the United States [25]. Variables collected include patient demographics, preoperative comorbidities, perioperative laboratory test values, perioperative adverse outcomes, and readmissions data for 30 days
Platelet Count Group Categorization and Adverse Event Plot Construction Platelet counts, along with mode-normalized relative risk of adverse events, were plotted on a histogram. Mode-normalized relative risk is reported as a moving average using 5 bins (bin width: 25,000/mL) to reduce noise in the data. A U-shaped curve was obtained (Fig. 1). Platelet counts above an “any adverse event”
1672
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
Fig. 1. Histogram of platelet count and plot of adverse event relative risk (expressed as a moving average) as a function of platelet count for TKA patients. Left y-axis refers to the histogram in the figure. The horizontal dashed lines denote relative risks of 1 and 1.5 as reference lines. Right y-axis refers to the line and scatter plot of adverse event data. Data are presented as a moving average and are normalized to the risk at the mode of the histogram giving us relative risks for different platelet counts. Vertical dashed lines and crosshatching denote the different platelet categories. Blue squares represent major adverse events, black circles represent minor adverse events, red diamond represents any adverse events. TKA, total knee arthroplasty.
relative risk value of 1.5 were classified as being either abnormally low or abnormally high. Figure 2 was constructed in the same way and shows risk of hospital readmissions, transfusions, and thromboembolic events (deep vein thrombosis or pulmonary embolism). These will be presented in the Results section. Based on this analysis, platelet counts 116,000/mL were defined as abnormally low platelets, platelet counts 492,000/mL were defined as abnormally high platelets, and the platelet counts in between these 2 cutoffs were defined as normal platelets.
Univariate Analysis An unadjusted comparison of patient demographic factors with normal, abnormally low, and abnormally high platelet counts was performed. Age, sex, BMI, functional status before surgery, ASA score, diabetes status, and smoker status were compared across normal, abnormally low, and abnormally high platelet count cohorts. Perioperative adverse outcomes were similarly assessed (any, major, minor adverse events and operative times and postoperative LOS). Chi-squared tests were used for categorical variables while 2tailed, 2-sample t-test using groups was used for continuous variables. Statistical significance was set at a ¼ 0.05 for both.
Multivariate Analysis Logistic regressions were performed for adverse events, readmissions, thromboembolic events, and transfusions for both the abnormally low and abnormally high platelet cohorts. The regression controlled for age, sex, BMI, ASA score, and functional status categories. Statistical significance was set at a ¼ 0.05 and 95% confidence intervals are reported. Statistical analysis was performed in Stata version 13.1 (StataCorp, LP, College Station, TX). Plots were constructed in MATLAB version 2017a (MathWorks, Inc, Natick, MA) and R: A Language and Environment for Statistical Computing version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria), running on RStudio version 1.0.153 (RStudio, Inc, Boston, MA).
Results Patient Population A total of 140,073 TKA patients were included in the study after selection based on inclusion and exclusion criteria. Platelet counts were separated into abnormally low and abnormally high categories using a relative risk threshold of 1.5 for any adverse event (Fig. 1). For TKA, these cutoffs were at 116,000/mL for abnormally low platelets and 492,000/mL for abnormally high platelets. It was noted that these TKA bounds for increased relative risk ratios were relatively dissimilar to reference values for platelet count in literature (150,000-400,000/mL). Based on the above-noted definitions, 137,915 (98.46%) fell into the normal platelet count group. Then, 1594 (1.14%) fell in the abnormally low platelet count group, and 564 (0.40%) fell in the abnormally high platelet count group. Comparison of Preoperative Characteristics Between Different Platelet Groups Compared to the normal platelet count group, the abnormally low platelet count cohort was older (P < .001), had more male patients (57.65% vs 37.37%, P < .001), and had a higher percentage of patients with dependent functional status (P ¼ .003), ASA scores 3 (P < .001), and diabetes (P < .001). Compared to the normal platelet count group, the abnormally high platelet count cohort had a higher percentage of younger patients (P < .001), females (78.55% vs 62.25%, P < .001), lower BMI values (P < .001), dependent functional status (P < .001), and smokers (P ¼ .004; Table 1). Adverse Outcome Analysis In the plot of adverse events vs platelet counts (Figs. 1 and 2), the rise in relative risk at both ends of the platelet count spectrum suggests an association between abnormal platelet counts and adverse events. The increase in relative risk also held in patients with more extreme values of lower/higher platelets.
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
1673
Fig. 2. Histogram of platelet count and plot of adverse events (in particular, hospital readmissions, transfusions, and thromboembolic events) as a function of platelet count for TKA patients. Similar to Figure 1, but the legend is different. Blue circles refer to readmissions, black triangles refer to thromboembolic events, and red asterisks refer to transfusions.
The univariate analyses of perioperative outcomes for abnormally high and low platelet groups are shown in Table 2. The abnormally low patient group was associated with a higher likelihood of any, major, and minor adverse events (all P < .001) relative to the normal platelet count cohort. The abnormally high patient group was associated with a higher likelihood of any and minor adverse events (both P < .001). Both the abnormal platelet cohorts were also associated with a higher risk of readmissions (abnormally low: P < .001, abnormally high: P ¼ .021). Furthermore, longer
hospital LOS was associated with both abnormally low and abnormally high platelet values (abnormally low: P < .001, abnormally high: P ¼ .010). Patients in the abnormally low platelet cohort also experienced longer operating times (P ¼ .016). Multivariate Analysis Based on the platelet cutoff values defined above, multivariate analysis for TKA revealed the abnormally low platelet cohort to be
Table 1 Demographics of Patients Undergoing TKA Focusing on Platelet Categories. Type
Normal Value
Platelet Count Total # Cases (N ¼ 140,073) Age 18-54 55-64 65-74 75 Sex Male Female BMI <25 25-30 30-35 >35 Functional status (before surgery) Independent Partially dependent Totally dependent ASA 1 2 3 4 Diabetes Insulin Noninsulin Smoker
Abnormal Value Low
Univariate
116-492 (1000s/mL)
116 (1000s/mL)
P Value
137,915 (98.46%)
1594 (1.14%)
Abnormal Value High
Univariate
492 (1000s/mL)
P Value
564 (0.40%) <.001
14,803 41,831 51,050 30,231
(10.73%) (30.33%) (37.02%) (21.92%)
122 448 549 475
(7.65%) (28.11%) (34.44%) (29.80%)
<.001 108 171 172 113
(19.15%) (30.32%) (30.50%) (20.04%)
<.001 52,067 (37.75%) 85,848 (62.25%)
919 (57.65%) 675 (42.35%)
13,915 37,594 39,300 47,106
165 419 478 532
<.001 121 (21.45%) 443 (78.55%) <.001
.539 (10.09%) (27.26%) (28.50%) (34.16%)
(10.35%) (26.29%) (29.99%) (33.38%)
126 160 145 133
(22.34%) (28.37%) (25.71%) (23.58%)
.003 136,141 (98.71%) 1719 (1.25%) 55 (0.04%)
1558 (97.74%) 35 (2.20%) 1 (0.00%)
.003 548 (97.16%) 16 (2.84%) 0 (0.00%)
<.001 2874 69,394 63,497 2150
(2.08%) (50.32%) (46.04%) (1.56%)
14 474 1031 75
(.88%) (29.74%) (64.68%) (4.71%)
.756 11 280 267 6
(1.95%) (49.65%) (47.34%) (1.06%)
<.001 5991 (4.34%) 18,511 (13.42%) 11,913 (8.64%)
148 (9.28%) 276 (17.31%) 168 (10.54%)
Italicized rows are the most common group. Bolding indicates statistical significance at P < .05. ASA, American Society of Anesthesiologists classification; TKA, total knee arthroplasty.
.007
.489 25 (4.43%) 66 (11.70%) 68 (12.06%)
.004
1674
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
Table 2 Platelet Range vs Postoperative Adverse Events for TKA. Type
Normal Value
Abnormal Value Low
Univariate
Abnormal Value High
Univariate
Platelet Count
116-492 (1000s/mL)
116 (1000s/mL)
P Value
492 (1000s/mL)
P Value
Total # Cases (N ¼ 140,073)
137,915 (98.46%)
1594 (1.14%)
Adverse event Major adverse event Death Sepsis/septic shock Unplanned intubation Ventilator >48 h Stroke Cardiac arrest Myocardial infarction Acute renal failure Thromboembolic event (PE, DVT) Wound infection Return to OR Readmission Minor adverse event Wound dehiscence Urinary tract infection Pneumonia Progressive renal insufficiency Transfusion
19,270 7565 180 344 202 95 113 105 290 76 1985 1113 1585 4664 13,383 261 1215 488 174 11,608
367 145 4 8 3 3 1 4 2 4 23 28 47 104 272 7 14 8 5 243
(13.97%) (5.49%) (0.13%) (0.25%) (0.15%) (0.07%) (0.08%) (0.08%) (0.21%) (0.06%) (1.44%) (0.81%) (1.15%) (3.38%) (9.70%) (0.19%) (0.88%) (0.35%) (0.13%) (8.42%)
(23.02%) (9.10%) (0.25%) (0.50%) (0.19%) (0.19%) (0.06%) (0.25%) (0.13%) (0.25%) (1.44%) (1.76%) (2.95%) (6.52%) (17.06%) (0.44%) (0.88%) (0.50%) (0.31%) (15.24%)
564 (0.40%) <.001 <.001
.998
<.001 <.001
<.001
123 39 2 1 1 0 1 0 1 0 12 4 9 29 97 3 3 3 1 88
(21.81%) (6.91%) (0.35%) (0.18%) (0.18%) (0.00%) (0.18%) (0.00%) (0.18%) (0.00%) (2.13%) (0.71%) (1.60%) (5.14%) (17.20%) (0.53%) (0.53%) (0.53%) (0.18%) (15.60%)
t-Test P Value Operating time (mins) Mean SD Length of stay (d) Mean SD
93.84 38.47
95.9 37.85
3.02 3.74
3.42 2.47
.016
<.001
<.001 .137
.171
.021 <.001
<.001 t-Test P Value
93.17 40.53
.654
3.24 2.22
.010
Bolding indicates statistical significance at P < .05. DVT, deep vein thrombosis; OR, operating room; PE, pulmonary embolism; SD, standard deviation; TKA, total knee arthroplasty.
associated with a higher likelihood of adverse events (any [odds ratio {OR} ¼ 1.66, P < .001], major [OR ¼ 1.46, P < .001], and minor [OR ¼ 1.79, P < .001] adverse events), a higher risk of readmission (OR ¼ 1.62, P < .001), and a higher risk of transfusions (OR ¼ 1.85, P < .001). The abnormally high platelet cohort was also associated with a higher likelihood of adverse events (any [OR ¼ 1.64, P < .001] and minor [OR ¼ 1.75, P < .001] adverse events), a higher risk of readmission (OR ¼ 1.66, P ¼ .008), as well as a higher risk of transfusions (OR ¼ 1.79, P < .001; Fig. 3). Discussion With rising rates of TKAs being performed every year, it is paramount to optimize outcomes following such surgical interventions [3]. Better risk stratification modeling and metrics not only help optimize patient care but also reduce the associated financial burden especially in the era of bundled care payments. Prior studies have examined associations between preoperative laboratory values and postoperative adverse events and had conflicting results. Furthermore, few studies have specifically examined the predictive value of preoperative platelet count on outcomes of TKA, which makes the present study unique. The purposes of the present study were to use the NSQIP database to determine the range of platelet counts of patients who have undergone TKA and to assess the association between platelet counts and postoperative adverse events and readmission while controlling for confounding factors. Notably, patients with an established diagnosis of any kind of neoplasm were excluded from the study. This extends to both solid tumors and hematologic malignancies. This serves to limit underlying causes of abnormal platelet counts. Although the study is not targeted toward determining the utility of
preoperative platelet counts, the data suggest the ability to risk stratify and change management in patients who have abnormal preoperative platelet counts. Standard definitions of thrombocytopenia and thrombocytosis are platelet count <150,000/mL and >450,000/mL, respectively [32]. In the present study, platelet counts were classified as either abnormally low or abnormally high if they were above an “any adverse event” relative risk value of 1.5. Based on this method, platelet counts 116,000/mL were defined as abnormally low, platelet counts 492,000/mL were defined as abnormally high, and platelet counts in between were defined as normal. It is of note that cutoffs from the present study are quite similar to commonly used values, differing by a maximum of only 42,000/mL. This highlights the strength of the method employed in the present study of basing platelet cutoffs on adverse event relative risk. Relative to those with normal platelet counts, those with lower platelet counts tended to be older, male, less functionally independent, of higher ASA, more likely diabetic, and smokers. In line with these findings, Glance et al [18] found that patients with severe thrombocytopenia tended to be older, male, and have more significant surgical histories and comorbidities. Analogously, those with higher platelet counts tended to be younger, female, of lower BMI, less functionally independent, and more likely smokers. Compared to patients with normal platelet counts, patients with abnormally low platelet counts had significantly longer operation times and hospital LOS. They also experienced almost 2-fold greater rates of any adverse event, major adverse event, readmission, minor adverse events, and blood transfusion. Patients with abnormally high platelet values experienced significantly longer LOS compared to patients with normal platelet counts. In addition, they experienced about 1.5-fold to 2-fold higher rates of any adverse event,
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676 Conditions and Outcomes
Odds Ratio
1675
P−Value
Odds Ratio [95% CI]
Abnormally low platelets Major adverse event Minor adverse event Any adverse event Hospital readmissions Thromboembolic event Transfusions
< .001 < .001 < .001 < .001 .948 < .001
1.46 [1.23, 1.74] 1.79 [1.56, 2.04] 1.66 [1.48, 1.88] 1.62 [1.33, 1.99] 0.99 [0.65, 1.49] 1.85 [1.60, 2.12]
Abnormally high platelets Major adverse event Minor adverse event Any adverse event Hospital readmissions Thromboembolic event Transfusions
.071 < .001 < .001 .008 .128 < .001
1.35 [0.97, 1.88] 1.75 [1.40, 2.19] 1.64 [1.34, 2.01] 1.66 [1.14, 2.42] 1.56 [0.88, 2.78] 1.79 [1.42, 2.25]
0
0.5
1
1.5
2
2.5
3
3.5
Odds Ratio
Fig. 3. Multivariate odds ratio of outcomes following TKA for abnormally low and abnormally high platelet categories. Contains details on multivariate analysis (logistic regression) on adverse event and readmission data for the different platelet categories and the forest plot. CI, confidence interval.
readmission, minor adverse events, and blood transfusion. From a system perspective, it is important to note that prior studies have indicated that frequent readmissions and prolonged LOS due to medical complications lead to increased cost burden [33,34]. When turning to multivariate analyses controlling for age, sex, BMI, ASA score, and functional status, the present study found that those with abnormally low platelet counts had higher odds of experiencing any adverse event (OR ¼ 1.66), major adverse event (OR ¼ 1.46), minor adverse event (OR ¼ 1.79), readmission (OR ¼ 1.62), and transfusion (OR ¼ 1.85). Analogously, with multivariate analysis controlling age, sex, BMI, ASA score, and functional status, the present study found that those with abnormally high platelet counts had higher odds of experiencing any adverse event (OR ¼ 1.64), minor adverse event (OR ¼ 1.75), and transfusion (OR ¼ 1.79). Prior studies have found that patients with moderate to severe thrombocytopenia were at greater risk of mortality and requiring transfusion 30 days following surgery [18]. Patients with thrombocytosis were at increased risk of requiring transfusion [16,18]. As shown in Figure 1, there is a U-shaped correlation between platelet count and adverse event. Abnormally low counts and abnormally high counts place patients at considerable risk of major, minor, and any adverse event. At less than 116,000 platelets/mL and greater than 492,000 platelets/mL, the relative risk of adverse event is at least 1.5 times greater than it is within normal range. Furthermore, at counts greater than 550,000 platelets/mL, the relative risk becomes 2 times greater than within normal limits. In other words, not only is there a higher risk with abnormally low and high platelet values, but the risk of adverse events increases as the counts become more abnormal. The present study was limited by factors that are common to most studies using large databases: it was limited to 30-day postoperative follow-up, and knee arthroplastyespecific outcome variables were not available for analysis. In addition, the reasons behind why some platelet counts are low/high are not present in the database, but there often is no definitive answer for such findings. The large sample size, high power, and generalizability of the present study are some of its greatest strengths. Additionally, the NSQIP database has been shown to be superior to others such as National Inpatient Sample in terms of robustness and accuracy [35,36]. NSQIP is highly powered and has quality data, which affords the present study an advantage over others that used administrative databases that are often associated with coding inconsistencies and data completion issues [35,37].
Conclusion The present study employed a large patient sample size and showed that elective TKA patients with abnormally high, as well as low, platelet counts are at increased risk of postoperative adverse outcomes. These findings suggest that preoperative laboratory values can in fact be useful in evaluating patients for TKA and could inform preoperative management to minimize postoperative risks and reduce associated costs. Focused attention needs to be paid to TKA patients with preoperative abnormal platelet counts for optimization and postoperative care. The data suggest that reversible causes of abnormal platelet counts might warrant a delay in performing the procedure while irreversible causes might warrant more careful follow-up to ameliorate adverse outcome risks.
References [1] Harris WH, Sledge CB. Total hip and total knee replacement (2). N Engl J Med 1990;323:801e7. [2] Liang MH, Cullen KE, Poss R. Primary total hip or knee replacement: evaluation of patients. Ann Intern Med 1982;97:735e9. [3] Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007;89:780e5. [4] Healthcare cost and utilization project. Agency for healthcare research and quality. In.: U.S. Department of Health & Human Services; [cited 2018/05/24]; Available at: http://hcupnet.ahrq.gov. [5] Bundled payments for care improvement (BPCI) initiative: general information. In.: Centers for medicare & medicaid services [updated 10/25/2018; cited 2018/05/23]; Available at: https://innovation.cms.gov/initiatives/bundled-pay ments/index.html. [6] Comprehensive care for joint replacement model. Centers for medicare & medicaid services. https://innovation.cms.gov/initiatives/cjr [accessed 22.05. 2018]. [7] Mechanic RE. Mandatory medicare bundled payment–is it ready for prime time? N Engl J Med 2015;373:1291e3. [8] Belmont Jr PJ, Goodman GP, Waterman BR, Bader JO, Schoenfeld AJ. Thirty-day postoperative complications and mortality following total knee arthroplasty: incidence and risk factors among a national sample of 15,321 patients. J Bone Joint Surg Am 2014;96:20e6. [9] Kreder HJ, Grosso P, Williams JI, Jaglal S, Axcell T, Wal EK, et al. Provider volume and other predictors of outcome after total knee arthroplasty: a population study in Ontario. Can J Surg 2003;46:15e22. [10] Winiarsky R, Barth P, Lotke P. Total knee arthroplasty in morbidly obese patients. J Bone Joint Surg Am 1998;80:1770e4. [11] Comprehensive care for joint replacement model. Centers for medicare & medicaid services. https://innovation.cms.gov/initiatives/cjr [accessed 30.11. 2017]. [12] Kerkhoffs GM, Servien E, Dunn W, Dahm D, Bramer JA, Haverkamp D. The influence of obesity on the complication rate and outcome of total knee
1676
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23] [24]
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676 arthroplasty: a meta-analysis and systematic literature review. J Bone Joint Surg Am 2012;94:1839e44. Namba RS, Inacio MC, Paxton EW. Risk factors associated with deep surgical site infections after primary total knee arthroplasty: an analysis of 56,216 knees. J Bone Joint Surg Am 2013;95:775e82. Namba RS, Paxton L, Fithian DC, Stone ML. Obesity and perioperative morbidity in total hip and total knee arthroplasty patients. J Arthroplasty 2005;20(7 Suppl 3):46e50. Peersman G, Laskin R, Davis J, Peterson M. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res 2001:15e23. Gibbs J, Cull W, Henderson W, Daley J, Hur K, Khuri SF. Preoperative serum albumin level as a predictor of operative mortality and morbidity: results from the National VA Surgical Risk Study. Arch Surg 1999;134:36e42. Saskin H, Duzyol C, Ozcan KS, Aksoy R, Idiz M. Preoperative platelet to lymphocyte ratio is associated with early morbidity and mortality after coronary artery bypass grafting. Heart Surg Forum 2015;18:E255e62. Glance LG, Blumberg N, Eaton MP, Lustik SJ, Osler TM, Wissler R, et al. Preoperative thrombocytopenia and postoperative outcomes after noncardiac surgery. Anesthesiology 2014;120:62e75. An TJ, Engstrom SM, Oelsner WK, Benvenuti MA, Polkowski GG, Schoenecker JG. Elevated d-dimer is not predictive of symptomatic deep venous thrombosis after total joint arthroplasty. J Arthroplasty 2016;31: 2269e72. 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;93:301e8. Kaplan EB, Sheiner LB, Boeckmann AJ, Roizen MF, Beal SL, Cohen SN, et al. The usefulness of preoperative laboratory screening. JAMA 1985;253: 3576e81. Narr BJ, Hansen TR, Warner MA. Preoperative laboratory screening in healthy Mayo patients: cost-effective elimination of tests and unchanged outcomes. Mayo Clin Proc 1991;66:155e9. Turnbull JM, Buck C. The value of preoperative screening investigations in otherwise healthy individuals. Arch Intern Med 1987;147:1101e5. Narr BJ, Warner ME, Schroeder DR, Warner MA. Outcomes of patients with no laboratory assessment before anesthesia and a surgical procedure. Mayo Clin Proc 1997;72:505e9.
[25] User Guide for the 2013 ACS-NSQIP Participant Use Data File. American College of Surgeons. 2014. Available at: https://www.facs.org/~/media/files/quality%20 programs/nsqip/2013_acs_nsqip_puf_user_guide.ashx. [accessed 20.05.2018]. [26] Bohl DD, Basques BA, Golinvaux NS, Miller CP, Baumgaertner MR, Grauer JN. Extramedullary compared with intramedullary implants for intertrochanteric hip fractures: thirty-day outcomes of 4432 procedures from the ACS NSQIP database. J Bone Joint Surg Am 2014;96:1871e7. [27] Bohl DD, Ondeck NT, Darrith B, Hannon CP, Fillingham YA, Della Valle CJ. Impact of operative time on adverse events following primary total joint arthroplasty. J Arthroplasty 2018;33:2256e2262.e4. [28] Bovonratwet P, Bohl DD, Malpani R, Nam D, Della Valle CJ, Grauer JN. Incidence, risk factors, and impact of Clostridium difficile colitis following primary total hip and knee arthroplasty. J Arthroplasty 2018;33:205e210.e1. [29] Donato DP, Kwok AC, Bishop MO, Presson AP, Agarwal JP. Unplanned readmission in outpatient hand surgery: an analysis of 23,613 patients in the NSQIP data set. Eplasty 2017;17:e36. [30] Qin C, Kim JY, Hsu WK. Impact of iinsulin dependence on lumbar surgery outcomes: an NSQIP analysis of 51,277 patients. Spine 2016;41:E687e93. [31] Basques BA, McLynn RP, Fice MP, Samuel AM, Lukasiewicz AM, Bohl DD, et al. Results of database studies in spine surgery can be influenced by missing data. Clin Orthop Relat Res 2017;475:2893e904. [32] Msaouel P, Lam AP, Gundabolu K, Chrysofakis G, Yu Y, Mantzaris I, et al. Abnormal platelet count is an independent predictor of mortality in the elderly and is influenced by ethnicity. Haematologica 2014;99:930e6. [33] Kapadia BH, McElroy MJ, Issa K, Johnson AJ, Bozic KJ, Mont MA. The economic impact of periprosthetic infections following total knee arthroplasty at a specialized tertiary-care center. J Arthroplasty 2014;29:929e32. [34] Schairer WW, Vail TP, Bozic KJ. What are the rates and causes of hospital readmission after total knee arthroplasty? Clin Orthop Relat Res 2014;472:181e7. [35] Kim L, Mabry C, Kimberg VS. Quality of benchmarks for assessment of care will influence outcome. J Craniofac Surg 2007;18:1504. [36] Weiss A, Anderson JE, Chang DC. Comparing the national surgical quality improvement program with the nationwide inpatient sample database. JAMA Surg 2015;150:815e6. [37] Faciszewski T, Broste SK, Fardon D. Quality of data regarding diagnoses of spinal disorders in administrative databases. A multicenter study. J Bone Joint Surg Am 1997;79:1481e8.
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Primary Arthroplasty
Abnormally High, as Well as Low, Preoperative Platelet Counts Correlate With Adverse Outcomes and Readmissions After Elective Total Knee Arthroplasty Rohil Malpani a, Monique S. Haynes, MPH a, Michael G. Clark b, Anoop R. Galivanche a, Patawut Bovonratwet a, Jonathan N. Grauer, MD a, * a b
Department of Orthopaedics and Rehabilitation, Yale School of Medicine, New Haven, Connecticut Vanderbilt University School of Medicine, Nashville, Tennessee
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 December 2018 Received in revised form 26 March 2019 Accepted 6 April 2019 Available online 18 April 2019
Background: Laboratory studies are routinely performed as a part of the preoperative workup for a total knee arthroplasty (TKA). The ramifications of abnormal preoperative platelet counts remain uncharacterized in large, multicenter patient populations. Methods: Patients who underwent elective primary TKA were identified in the 2011-2015 National Surgical Quality Improvement Program database. Risk of 30-day postoperative complications was calculated as a function of preoperative platelet counts. Patients were characterized as having a normal platelet count, abnormally low platelet count, and abnormally high platelet count based on relative risk calculations. Univariate and multivariate analyses were performed to associate abnormal platelet counts with patient demographics, operative variables, 30-day postoperative complications, and readmissions. Results: In total, 140,073 patients who underwent elective TKA were identified. Using the relative risk threshold of 1.5 for any adverse event, abnormally low and abnormally high platelet count thresholds were set at 116,000/mL and 492,000/mL, respectively. Multivariate analyses revealed low platelet counts to be associated with higher rates of any, major, and minor adverse events and longer length of stay. Analogously, high platelet counts were associated with higher rates of any and minor adverse events and longer length of stay. Conclusion: The present study employed a large patient sample size and showed that elective TKA patients with abnormally high, as well as low, platelet counts are at increased risk of postoperative adverse outcomes. Focused attention needs to be paid to TKA patients with preoperative abnormal platelet counts for optimization and postoperative care. Level of Evidence: Level III, retrospective comparative study. © 2019 Elsevier Inc. All rights reserved.
Keywords: total knee arthroplasty the American College of Surgeons National Surgical Quality Improvement Program platelet count surgical outcomes postoperative adverse events readmission
Total knee arthroplasty (TKA) is a common and reliable surgical procedure for patients with degenerative knee conditions and functional impairment refractory to conservative treatment [1,2]. Disclaimer: Financial remuneration to authors and family members related to the subject of this article: None. One or more of the authors of this paper have disclosed potential or pertinent conflicts of interest, which may include receipt of payment, either direct or indirect, institutional support, or association with an entity in the biomedical field which may be perceived to have potential conflict of interest with this work. For full disclosure statements refer to https://doi.org/10.1016/j.arth.2019.04.012. * Reprint requests: Jonathan N. Grauer, MD, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, 47 College Street, New Haven, CT 06520. https://doi.org/10.1016/j.arth.2019.04.012 0883-5403/© 2019 Elsevier Inc. All rights reserved.
More than 600,000 TKA procedures are performed annually in the United States, and regression models predict demand for primary TKA will exceed 3 million procedures annually by 2030 [3,4]. Optimizing outcomes after TKA is critical given the increase in surgical volume. Rising demand for TKA could place a financial burden on patients and hospital systems [3]. In the United States, the Bundled Payments for Care Improvement and Comprehensive Care for Joint Replacement models have been proposed to increase quality and coordination of care while reducing costs to Medicare [5,6]. Under these models, hospitals have a financial incentive to increase quality of care (ie, improve outcomes) and minimize cost (ie, reduce the frequency of surgical complications) [7]. These motivations
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
heighten the need for preoperative risk stratification models which could be used by providers to identify higher-risk surgeries and take steps to mitigate risk. Several qualitative and quantitative metrics for preoperative risk stratification have been evaluated. Patient age, for example, has been associated with increased mortality [8] and incidence of postoperative complications [8,9]. Similarly, high body mass index (BMI) has been associated with increased incidence of perioperative and postoperative complications [8,10], increased postoperative infection risk [11e15], and increased rates of revision surgery [12]. The presence of preexisting comorbidities, such as diabetes and cardiac disease, has been associated with increased risk of mortality [8], surgical complications [8,9], and infection [13,15]. Overall medical status also appears to be associated with surgical outcomes: American Society of Anesthesiologists (ASA) classifications 3 have been associated with higher incidence of postoperative complications [8] and infection [13]. Preoperative laboratory studies, such as chemistry panels, complete blood counts, and coagulation studies, are prime candidates for use in risk stratification models due to their ubiquity in the standard presurgical workup. Previous work has shown that low serum albumin levels are a strong predictor of surgical mortality and morbidity (especially sepsis and major infections) [16]. In cardiac surgery, increased platelet-to-lymphocyte ratio was shown to be associated with mortality and morbidity [17]. In noncardiac surgery, abnormal preoperative platelet counts have been associated with a higher incidence of blood transfusion and higher risk of death [18]. Other analyses have not found utility in certain preoperative laboratory studies. Serum D-dimer, for example, was shown to be an ineffective screening test for postoperative deep vein thrombosis following TKA [19]. Abnormal electrolyte, hemoglobin, creatinine, and glucose values, as well as thrombocytopenia, were not predictive of postoperative adverse outcomes in a study of elderly patients undergoing noncardiac surgery [20]. Other studies have suggested that routine preoperative laboratory screening tests contribute little to patient care and decision-making [21e23] and that not obtaining them does not adversely affect patient safety [24]. Many of the aforementioned positive and negative studies suffer from low cohort size (<1000) [17,19,20] or are not specific to TKA [16e18,20e24]. Consequently, the usefulness and clinical implications of abnormal preoperative platelet counts in patients undergoing primary TKA have not been fully established. As a laboratory test frequently ordered preoperatively, there is a need to aid in the interpretation of platelet count values before surgery, as well as establish relationships between preoperative values and postoperative events. To address this, the present study queried the National Surgical Quality Improvement Program (NSQIP) database to identify >100,000 patients who underwent TKA between 2011 and 2015. These data were interrogated to quantitatively characterize the range of platelet counts for patients undergoing primary TKA and to investigate associations between abnormal preoperative platelet counts and postoperative complications.
1671
following the procedure, regardless of discharge status. This database has been used extensively in various surgical specialty research, and its use in the orthopedic field has been extensive, dramatically increasing in recent years with higher-volume data [26e30]. Our institutional review board granted an exemption for studies using this dataset. Inclusion/Exclusion Criteria Patients who underwent elective primary TKA between 2011 and 2015 were identified using the Current Procedural Terminology code 27447. Procedures before 2011 did not include 30-day readmission data and were therefore not considered for the present study. Exclusion criteria included procedures performed for fracture, neoplasm, and infection. Patients who had a revision TKA (Current Procedural Terminology code 27486/27487) or who underwent nonelective TKA were also excluded. After exclusion, based on these criteria, a small proportion (<1%) of patients had missing data for vital demographic attributes and were excluded. An additional 5.25% of the patients had missing preoperative platelet count values and were also excluded. Of note, it has been shown in the orthopedic literature that variables with <10% missing data do not appreciably skew results of complete case analysis [31].
Materials and Methods
Patient Preoperative Characteristics and Perioperative Complications Age, sex, height, weight, functional status before surgery, ASA classification, smoking status, diabetes mellitus status, and preoperative platelet count were queried and directly abstracted from the dataset. Using height and weight data, BMI (in kg/m2) was calculated. ASA was used as a marker of comorbidity. Operating time and hospital length of stay (LOS) were abstracted from the database. Operating time (incision to closure) was measured in minutes while LOS (admission to discharge) was measured in days. The maximum value of the LOS was limited to 30 days. The NSQIP database contains data describing adverse outcomes from the 30-day follow-up period. Individual adverse events and aggregated any, major, and minor adverse events were assessed. The occurrence of a major adverse event was defined as the occurrence of at least one of the following: death, return to operating room, sepsis/septic shock, unplanned intubation, ventilator use >48 hours, stroke/cerebrovascular accident, cardiac arrest, acute renal failure, thromboembolic event (defined as pulmonary embolism or deep vein thrombosis), wound infection, return to the operating room, and 30-day readmissions. The occurrence of a minor adverse event was defined as the occurrence of any of the following: wound dehiscence, urinary tract infection, pneumonia, progressive renal insufficiency, and transfusions (a transfusion is defined in NSQIP as occurrence of transfusion of at least 1 unit of packed red blood cells starting from time of onset of the procedure up to 72 hours postoperatively, including hanging blood from the operating room that is finished outside the operating room, except NSQIP data from 2011 that did not collect intraoperative blood transfusion data). Any adverse event was defined as the occurrence of either a major or minor adverse event.
Database and Patient Population
Statistical Analysis
Database The NSQIP database contains over 300 variables for each surgical case from more than 500 participating healthcare institutions, the majority of which are based in the United States [25]. Variables collected include patient demographics, preoperative comorbidities, perioperative laboratory test values, perioperative adverse outcomes, and readmissions data for 30 days
Platelet Count Group Categorization and Adverse Event Plot Construction Platelet counts, along with mode-normalized relative risk of adverse events, were plotted on a histogram. Mode-normalized relative risk is reported as a moving average using 5 bins (bin width: 25,000/mL) to reduce noise in the data. A U-shaped curve was obtained (Fig. 1). Platelet counts above an “any adverse event”
1672
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
Fig. 1. Histogram of platelet count and plot of adverse event relative risk (expressed as a moving average) as a function of platelet count for TKA patients. Left y-axis refers to the histogram in the figure. The horizontal dashed lines denote relative risks of 1 and 1.5 as reference lines. Right y-axis refers to the line and scatter plot of adverse event data. Data are presented as a moving average and are normalized to the risk at the mode of the histogram giving us relative risks for different platelet counts. Vertical dashed lines and crosshatching denote the different platelet categories. Blue squares represent major adverse events, black circles represent minor adverse events, red diamond represents any adverse events. TKA, total knee arthroplasty.
relative risk value of 1.5 were classified as being either abnormally low or abnormally high. Figure 2 was constructed in the same way and shows risk of hospital readmissions, transfusions, and thromboembolic events (deep vein thrombosis or pulmonary embolism). These will be presented in the Results section. Based on this analysis, platelet counts 116,000/mL were defined as abnormally low platelets, platelet counts 492,000/mL were defined as abnormally high platelets, and the platelet counts in between these 2 cutoffs were defined as normal platelets.
Univariate Analysis An unadjusted comparison of patient demographic factors with normal, abnormally low, and abnormally high platelet counts was performed. Age, sex, BMI, functional status before surgery, ASA score, diabetes status, and smoker status were compared across normal, abnormally low, and abnormally high platelet count cohorts. Perioperative adverse outcomes were similarly assessed (any, major, minor adverse events and operative times and postoperative LOS). Chi-squared tests were used for categorical variables while 2tailed, 2-sample t-test using groups was used for continuous variables. Statistical significance was set at a ¼ 0.05 for both.
Multivariate Analysis Logistic regressions were performed for adverse events, readmissions, thromboembolic events, and transfusions for both the abnormally low and abnormally high platelet cohorts. The regression controlled for age, sex, BMI, ASA score, and functional status categories. Statistical significance was set at a ¼ 0.05 and 95% confidence intervals are reported. Statistical analysis was performed in Stata version 13.1 (StataCorp, LP, College Station, TX). Plots were constructed in MATLAB version 2017a (MathWorks, Inc, Natick, MA) and R: A Language and Environment for Statistical Computing version 3.2.3 (R Foundation for Statistical Computing, Vienna, Austria), running on RStudio version 1.0.153 (RStudio, Inc, Boston, MA).
Results Patient Population A total of 140,073 TKA patients were included in the study after selection based on inclusion and exclusion criteria. Platelet counts were separated into abnormally low and abnormally high categories using a relative risk threshold of 1.5 for any adverse event (Fig. 1). For TKA, these cutoffs were at 116,000/mL for abnormally low platelets and 492,000/mL for abnormally high platelets. It was noted that these TKA bounds for increased relative risk ratios were relatively dissimilar to reference values for platelet count in literature (150,000-400,000/mL). Based on the above-noted definitions, 137,915 (98.46%) fell into the normal platelet count group. Then, 1594 (1.14%) fell in the abnormally low platelet count group, and 564 (0.40%) fell in the abnormally high platelet count group. Comparison of Preoperative Characteristics Between Different Platelet Groups Compared to the normal platelet count group, the abnormally low platelet count cohort was older (P < .001), had more male patients (57.65% vs 37.37%, P < .001), and had a higher percentage of patients with dependent functional status (P ¼ .003), ASA scores 3 (P < .001), and diabetes (P < .001). Compared to the normal platelet count group, the abnormally high platelet count cohort had a higher percentage of younger patients (P < .001), females (78.55% vs 62.25%, P < .001), lower BMI values (P < .001), dependent functional status (P < .001), and smokers (P ¼ .004; Table 1). Adverse Outcome Analysis In the plot of adverse events vs platelet counts (Figs. 1 and 2), the rise in relative risk at both ends of the platelet count spectrum suggests an association between abnormal platelet counts and adverse events. The increase in relative risk also held in patients with more extreme values of lower/higher platelets.
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
1673
Fig. 2. Histogram of platelet count and plot of adverse events (in particular, hospital readmissions, transfusions, and thromboembolic events) as a function of platelet count for TKA patients. Similar to Figure 1, but the legend is different. Blue circles refer to readmissions, black triangles refer to thromboembolic events, and red asterisks refer to transfusions.
The univariate analyses of perioperative outcomes for abnormally high and low platelet groups are shown in Table 2. The abnormally low patient group was associated with a higher likelihood of any, major, and minor adverse events (all P < .001) relative to the normal platelet count cohort. The abnormally high patient group was associated with a higher likelihood of any and minor adverse events (both P < .001). Both the abnormal platelet cohorts were also associated with a higher risk of readmissions (abnormally low: P < .001, abnormally high: P ¼ .021). Furthermore, longer
hospital LOS was associated with both abnormally low and abnormally high platelet values (abnormally low: P < .001, abnormally high: P ¼ .010). Patients in the abnormally low platelet cohort also experienced longer operating times (P ¼ .016). Multivariate Analysis Based on the platelet cutoff values defined above, multivariate analysis for TKA revealed the abnormally low platelet cohort to be
Table 1 Demographics of Patients Undergoing TKA Focusing on Platelet Categories. Type
Normal Value
Platelet Count Total # Cases (N ¼ 140,073) Age 18-54 55-64 65-74 75 Sex Male Female BMI <25 25-30 30-35 >35 Functional status (before surgery) Independent Partially dependent Totally dependent ASA 1 2 3 4 Diabetes Insulin Noninsulin Smoker
Abnormal Value Low
Univariate
116-492 (1000s/mL)
116 (1000s/mL)
P Value
137,915 (98.46%)
1594 (1.14%)
Abnormal Value High
Univariate
492 (1000s/mL)
P Value
564 (0.40%) <.001
14,803 41,831 51,050 30,231
(10.73%) (30.33%) (37.02%) (21.92%)
122 448 549 475
(7.65%) (28.11%) (34.44%) (29.80%)
<.001 108 171 172 113
(19.15%) (30.32%) (30.50%) (20.04%)
<.001 52,067 (37.75%) 85,848 (62.25%)
919 (57.65%) 675 (42.35%)
13,915 37,594 39,300 47,106
165 419 478 532
<.001 121 (21.45%) 443 (78.55%) <.001
.539 (10.09%) (27.26%) (28.50%) (34.16%)
(10.35%) (26.29%) (29.99%) (33.38%)
126 160 145 133
(22.34%) (28.37%) (25.71%) (23.58%)
.003 136,141 (98.71%) 1719 (1.25%) 55 (0.04%)
1558 (97.74%) 35 (2.20%) 1 (0.00%)
.003 548 (97.16%) 16 (2.84%) 0 (0.00%)
<.001 2874 69,394 63,497 2150
(2.08%) (50.32%) (46.04%) (1.56%)
14 474 1031 75
(.88%) (29.74%) (64.68%) (4.71%)
.756 11 280 267 6
(1.95%) (49.65%) (47.34%) (1.06%)
<.001 5991 (4.34%) 18,511 (13.42%) 11,913 (8.64%)
148 (9.28%) 276 (17.31%) 168 (10.54%)
Italicized rows are the most common group. Bolding indicates statistical significance at P < .05. ASA, American Society of Anesthesiologists classification; TKA, total knee arthroplasty.
.007
.489 25 (4.43%) 66 (11.70%) 68 (12.06%)
.004
1674
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676
Table 2 Platelet Range vs Postoperative Adverse Events for TKA. Type
Normal Value
Abnormal Value Low
Univariate
Abnormal Value High
Univariate
Platelet Count
116-492 (1000s/mL)
116 (1000s/mL)
P Value
492 (1000s/mL)
P Value
Total # Cases (N ¼ 140,073)
137,915 (98.46%)
1594 (1.14%)
Adverse event Major adverse event Death Sepsis/septic shock Unplanned intubation Ventilator >48 h Stroke Cardiac arrest Myocardial infarction Acute renal failure Thromboembolic event (PE, DVT) Wound infection Return to OR Readmission Minor adverse event Wound dehiscence Urinary tract infection Pneumonia Progressive renal insufficiency Transfusion
19,270 7565 180 344 202 95 113 105 290 76 1985 1113 1585 4664 13,383 261 1215 488 174 11,608
367 145 4 8 3 3 1 4 2 4 23 28 47 104 272 7 14 8 5 243
(13.97%) (5.49%) (0.13%) (0.25%) (0.15%) (0.07%) (0.08%) (0.08%) (0.21%) (0.06%) (1.44%) (0.81%) (1.15%) (3.38%) (9.70%) (0.19%) (0.88%) (0.35%) (0.13%) (8.42%)
(23.02%) (9.10%) (0.25%) (0.50%) (0.19%) (0.19%) (0.06%) (0.25%) (0.13%) (0.25%) (1.44%) (1.76%) (2.95%) (6.52%) (17.06%) (0.44%) (0.88%) (0.50%) (0.31%) (15.24%)
564 (0.40%) <.001 <.001
.998
<.001 <.001
<.001
123 39 2 1 1 0 1 0 1 0 12 4 9 29 97 3 3 3 1 88
(21.81%) (6.91%) (0.35%) (0.18%) (0.18%) (0.00%) (0.18%) (0.00%) (0.18%) (0.00%) (2.13%) (0.71%) (1.60%) (5.14%) (17.20%) (0.53%) (0.53%) (0.53%) (0.18%) (15.60%)
t-Test P Value Operating time (mins) Mean SD Length of stay (d) Mean SD
93.84 38.47
95.9 37.85
3.02 3.74
3.42 2.47
.016
<.001
<.001 .137
.171
.021 <.001
<.001 t-Test P Value
93.17 40.53
.654
3.24 2.22
.010
Bolding indicates statistical significance at P < .05. DVT, deep vein thrombosis; OR, operating room; PE, pulmonary embolism; SD, standard deviation; TKA, total knee arthroplasty.
associated with a higher likelihood of adverse events (any [odds ratio {OR} ¼ 1.66, P < .001], major [OR ¼ 1.46, P < .001], and minor [OR ¼ 1.79, P < .001] adverse events), a higher risk of readmission (OR ¼ 1.62, P < .001), and a higher risk of transfusions (OR ¼ 1.85, P < .001). The abnormally high platelet cohort was also associated with a higher likelihood of adverse events (any [OR ¼ 1.64, P < .001] and minor [OR ¼ 1.75, P < .001] adverse events), a higher risk of readmission (OR ¼ 1.66, P ¼ .008), as well as a higher risk of transfusions (OR ¼ 1.79, P < .001; Fig. 3). Discussion With rising rates of TKAs being performed every year, it is paramount to optimize outcomes following such surgical interventions [3]. Better risk stratification modeling and metrics not only help optimize patient care but also reduce the associated financial burden especially in the era of bundled care payments. Prior studies have examined associations between preoperative laboratory values and postoperative adverse events and had conflicting results. Furthermore, few studies have specifically examined the predictive value of preoperative platelet count on outcomes of TKA, which makes the present study unique. The purposes of the present study were to use the NSQIP database to determine the range of platelet counts of patients who have undergone TKA and to assess the association between platelet counts and postoperative adverse events and readmission while controlling for confounding factors. Notably, patients with an established diagnosis of any kind of neoplasm were excluded from the study. This extends to both solid tumors and hematologic malignancies. This serves to limit underlying causes of abnormal platelet counts. Although the study is not targeted toward determining the utility of
preoperative platelet counts, the data suggest the ability to risk stratify and change management in patients who have abnormal preoperative platelet counts. Standard definitions of thrombocytopenia and thrombocytosis are platelet count <150,000/mL and >450,000/mL, respectively [32]. In the present study, platelet counts were classified as either abnormally low or abnormally high if they were above an “any adverse event” relative risk value of 1.5. Based on this method, platelet counts 116,000/mL were defined as abnormally low, platelet counts 492,000/mL were defined as abnormally high, and platelet counts in between were defined as normal. It is of note that cutoffs from the present study are quite similar to commonly used values, differing by a maximum of only 42,000/mL. This highlights the strength of the method employed in the present study of basing platelet cutoffs on adverse event relative risk. Relative to those with normal platelet counts, those with lower platelet counts tended to be older, male, less functionally independent, of higher ASA, more likely diabetic, and smokers. In line with these findings, Glance et al [18] found that patients with severe thrombocytopenia tended to be older, male, and have more significant surgical histories and comorbidities. Analogously, those with higher platelet counts tended to be younger, female, of lower BMI, less functionally independent, and more likely smokers. Compared to patients with normal platelet counts, patients with abnormally low platelet counts had significantly longer operation times and hospital LOS. They also experienced almost 2-fold greater rates of any adverse event, major adverse event, readmission, minor adverse events, and blood transfusion. Patients with abnormally high platelet values experienced significantly longer LOS compared to patients with normal platelet counts. In addition, they experienced about 1.5-fold to 2-fold higher rates of any adverse event,
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676 Conditions and Outcomes
Odds Ratio
1675
P−Value
Odds Ratio [95% CI]
Abnormally low platelets Major adverse event Minor adverse event Any adverse event Hospital readmissions Thromboembolic event Transfusions
< .001 < .001 < .001 < .001 .948 < .001
1.46 [1.23, 1.74] 1.79 [1.56, 2.04] 1.66 [1.48, 1.88] 1.62 [1.33, 1.99] 0.99 [0.65, 1.49] 1.85 [1.60, 2.12]
Abnormally high platelets Major adverse event Minor adverse event Any adverse event Hospital readmissions Thromboembolic event Transfusions
.071 < .001 < .001 .008 .128 < .001
1.35 [0.97, 1.88] 1.75 [1.40, 2.19] 1.64 [1.34, 2.01] 1.66 [1.14, 2.42] 1.56 [0.88, 2.78] 1.79 [1.42, 2.25]
0
0.5
1
1.5
2
2.5
3
3.5
Odds Ratio
Fig. 3. Multivariate odds ratio of outcomes following TKA for abnormally low and abnormally high platelet categories. Contains details on multivariate analysis (logistic regression) on adverse event and readmission data for the different platelet categories and the forest plot. CI, confidence interval.
readmission, minor adverse events, and blood transfusion. From a system perspective, it is important to note that prior studies have indicated that frequent readmissions and prolonged LOS due to medical complications lead to increased cost burden [33,34]. When turning to multivariate analyses controlling for age, sex, BMI, ASA score, and functional status, the present study found that those with abnormally low platelet counts had higher odds of experiencing any adverse event (OR ¼ 1.66), major adverse event (OR ¼ 1.46), minor adverse event (OR ¼ 1.79), readmission (OR ¼ 1.62), and transfusion (OR ¼ 1.85). Analogously, with multivariate analysis controlling age, sex, BMI, ASA score, and functional status, the present study found that those with abnormally high platelet counts had higher odds of experiencing any adverse event (OR ¼ 1.64), minor adverse event (OR ¼ 1.75), and transfusion (OR ¼ 1.79). Prior studies have found that patients with moderate to severe thrombocytopenia were at greater risk of mortality and requiring transfusion 30 days following surgery [18]. Patients with thrombocytosis were at increased risk of requiring transfusion [16,18]. As shown in Figure 1, there is a U-shaped correlation between platelet count and adverse event. Abnormally low counts and abnormally high counts place patients at considerable risk of major, minor, and any adverse event. At less than 116,000 platelets/mL and greater than 492,000 platelets/mL, the relative risk of adverse event is at least 1.5 times greater than it is within normal range. Furthermore, at counts greater than 550,000 platelets/mL, the relative risk becomes 2 times greater than within normal limits. In other words, not only is there a higher risk with abnormally low and high platelet values, but the risk of adverse events increases as the counts become more abnormal. The present study was limited by factors that are common to most studies using large databases: it was limited to 30-day postoperative follow-up, and knee arthroplastyespecific outcome variables were not available for analysis. In addition, the reasons behind why some platelet counts are low/high are not present in the database, but there often is no definitive answer for such findings. The large sample size, high power, and generalizability of the present study are some of its greatest strengths. Additionally, the NSQIP database has been shown to be superior to others such as National Inpatient Sample in terms of robustness and accuracy [35,36]. NSQIP is highly powered and has quality data, which affords the present study an advantage over others that used administrative databases that are often associated with coding inconsistencies and data completion issues [35,37].
Conclusion The present study employed a large patient sample size and showed that elective TKA patients with abnormally high, as well as low, platelet counts are at increased risk of postoperative adverse outcomes. These findings suggest that preoperative laboratory values can in fact be useful in evaluating patients for TKA and could inform preoperative management to minimize postoperative risks and reduce associated costs. Focused attention needs to be paid to TKA patients with preoperative abnormal platelet counts for optimization and postoperative care. The data suggest that reversible causes of abnormal platelet counts might warrant a delay in performing the procedure while irreversible causes might warrant more careful follow-up to ameliorate adverse outcome risks.
References [1] Harris WH, Sledge CB. Total hip and total knee replacement (2). N Engl J Med 1990;323:801e7. [2] Liang MH, Cullen KE, Poss R. Primary total hip or knee replacement: evaluation of patients. Ann Intern Med 1982;97:735e9. [3] Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007;89:780e5. [4] Healthcare cost and utilization project. Agency for healthcare research and quality. In.: U.S. Department of Health & Human Services; [cited 2018/05/24]; Available at: http://hcupnet.ahrq.gov. [5] Bundled payments for care improvement (BPCI) initiative: general information. In.: Centers for medicare & medicaid services [updated 10/25/2018; cited 2018/05/23]; Available at: https://innovation.cms.gov/initiatives/bundled-pay ments/index.html. [6] Comprehensive care for joint replacement model. Centers for medicare & medicaid services. https://innovation.cms.gov/initiatives/cjr [accessed 22.05. 2018]. [7] Mechanic RE. Mandatory medicare bundled payment–is it ready for prime time? N Engl J Med 2015;373:1291e3. [8] Belmont Jr PJ, Goodman GP, Waterman BR, Bader JO, Schoenfeld AJ. Thirty-day postoperative complications and mortality following total knee arthroplasty: incidence and risk factors among a national sample of 15,321 patients. J Bone Joint Surg Am 2014;96:20e6. [9] Kreder HJ, Grosso P, Williams JI, Jaglal S, Axcell T, Wal EK, et al. Provider volume and other predictors of outcome after total knee arthroplasty: a population study in Ontario. Can J Surg 2003;46:15e22. [10] Winiarsky R, Barth P, Lotke P. Total knee arthroplasty in morbidly obese patients. J Bone Joint Surg Am 1998;80:1770e4. [11] Comprehensive care for joint replacement model. Centers for medicare & medicaid services. https://innovation.cms.gov/initiatives/cjr [accessed 30.11. 2017]. [12] Kerkhoffs GM, Servien E, Dunn W, Dahm D, Bramer JA, Haverkamp D. The influence of obesity on the complication rate and outcome of total knee
1676
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23] [24]
R. Malpani et al. / The Journal of Arthroplasty 34 (2019) 1670e1676 arthroplasty: a meta-analysis and systematic literature review. J Bone Joint Surg Am 2012;94:1839e44. Namba RS, Inacio MC, Paxton EW. Risk factors associated with deep surgical site infections after primary total knee arthroplasty: an analysis of 56,216 knees. J Bone Joint Surg Am 2013;95:775e82. Namba RS, Paxton L, Fithian DC, Stone ML. Obesity and perioperative morbidity in total hip and total knee arthroplasty patients. J Arthroplasty 2005;20(7 Suppl 3):46e50. Peersman G, Laskin R, Davis J, Peterson M. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res 2001:15e23. Gibbs J, Cull W, Henderson W, Daley J, Hur K, Khuri SF. Preoperative serum albumin level as a predictor of operative mortality and morbidity: results from the National VA Surgical Risk Study. Arch Surg 1999;134:36e42. Saskin H, Duzyol C, Ozcan KS, Aksoy R, Idiz M. Preoperative platelet to lymphocyte ratio is associated with early morbidity and mortality after coronary artery bypass grafting. Heart Surg Forum 2015;18:E255e62. Glance LG, Blumberg N, Eaton MP, Lustik SJ, Osler TM, Wissler R, et al. Preoperative thrombocytopenia and postoperative outcomes after noncardiac surgery. Anesthesiology 2014;120:62e75. An TJ, Engstrom SM, Oelsner WK, Benvenuti MA, Polkowski GG, Schoenecker JG. Elevated d-dimer is not predictive of symptomatic deep venous thrombosis after total joint arthroplasty. J Arthroplasty 2016;31: 2269e72. 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;93:301e8. Kaplan EB, Sheiner LB, Boeckmann AJ, Roizen MF, Beal SL, Cohen SN, et al. The usefulness of preoperative laboratory screening. JAMA 1985;253: 3576e81. Narr BJ, Hansen TR, Warner MA. Preoperative laboratory screening in healthy Mayo patients: cost-effective elimination of tests and unchanged outcomes. Mayo Clin Proc 1991;66:155e9. Turnbull JM, Buck C. The value of preoperative screening investigations in otherwise healthy individuals. Arch Intern Med 1987;147:1101e5. Narr BJ, Warner ME, Schroeder DR, Warner MA. Outcomes of patients with no laboratory assessment before anesthesia and a surgical procedure. Mayo Clin Proc 1997;72:505e9.
[25] User Guide for the 2013 ACS-NSQIP Participant Use Data File. American College of Surgeons. 2014. Available at: https://www.facs.org/~/media/files/quality%20 programs/nsqip/2013_acs_nsqip_puf_user_guide.ashx. [accessed 20.05.2018]. [26] Bohl DD, Basques BA, Golinvaux NS, Miller CP, Baumgaertner MR, Grauer JN. Extramedullary compared with intramedullary implants for intertrochanteric hip fractures: thirty-day outcomes of 4432 procedures from the ACS NSQIP database. J Bone Joint Surg Am 2014;96:1871e7. [27] Bohl DD, Ondeck NT, Darrith B, Hannon CP, Fillingham YA, Della Valle CJ. Impact of operative time on adverse events following primary total joint arthroplasty. J Arthroplasty 2018;33:2256e2262.e4. [28] Bovonratwet P, Bohl DD, Malpani R, Nam D, Della Valle CJ, Grauer JN. Incidence, risk factors, and impact of Clostridium difficile colitis following primary total hip and knee arthroplasty. J Arthroplasty 2018;33:205e210.e1. [29] Donato DP, Kwok AC, Bishop MO, Presson AP, Agarwal JP. Unplanned readmission in outpatient hand surgery: an analysis of 23,613 patients in the NSQIP data set. Eplasty 2017;17:e36. [30] Qin C, Kim JY, Hsu WK. Impact of iinsulin dependence on lumbar surgery outcomes: an NSQIP analysis of 51,277 patients. Spine 2016;41:E687e93. [31] Basques BA, McLynn RP, Fice MP, Samuel AM, Lukasiewicz AM, Bohl DD, et al. Results of database studies in spine surgery can be influenced by missing data. Clin Orthop Relat Res 2017;475:2893e904. [32] Msaouel P, Lam AP, Gundabolu K, Chrysofakis G, Yu Y, Mantzaris I, et al. Abnormal platelet count is an independent predictor of mortality in the elderly and is influenced by ethnicity. Haematologica 2014;99:930e6. [33] Kapadia BH, McElroy MJ, Issa K, Johnson AJ, Bozic KJ, Mont MA. The economic impact of periprosthetic infections following total knee arthroplasty at a specialized tertiary-care center. J Arthroplasty 2014;29:929e32. [34] Schairer WW, Vail TP, Bozic KJ. What are the rates and causes of hospital readmission after total knee arthroplasty? Clin Orthop Relat Res 2014;472:181e7. [35] Kim L, Mabry C, Kimberg VS. Quality of benchmarks for assessment of care will influence outcome. J Craniofac Surg 2007;18:1504. [36] Weiss A, Anderson JE, Chang DC. Comparing the national surgical quality improvement program with the nationwide inpatient sample database. JAMA Surg 2015;150:815e6. [37] Faciszewski T, Broste SK, Fardon D. Quality of data regarding diagnoses of spinal disorders in administrative databases. A multicenter study. J Bone Joint Surg Am 1997;79:1481e8.