Preoperative Hemoglobin Levels, Characteristics, and Resource Utilization in Total Knee Replacement Patients with An Anemia Diagnosis Versus without An Anemia Diagnosis: An Electronic Medical Record Analysis

Abstract

Abstract 2089 Purpose:This retrospective analysis describes preoperative hemoglobin (Hb) levels, characteristics, and resource use in total knee replacement (TKR) patients who had an anemia diagnosis as compared to those without an anemia diagnosis. Methods:Electronic medical records (EMR) from a large US integrated health delivery system were analyzed for the period 01/2004 to 09/2010. Adult patients with a TKR surgery record and ≥1 Hb measurement were studied. Patients with a hip or knee revision before or during surgery, with bilateral surgery, or with an emergency room (ER) visit on the surgery admission date were excluded. Patients were stratified into two groups based upon the presence or absence of an anemia diagnosis (ICD-9 codes 280.xx-285.xx) in the 90 days before or day of surgery. Hemoglobin levels measured in the 45 days prior to but excluding the day of surgery were analyzed. For patients with multiple Hb measurements, the earliest observed Hb measurement (defined as the measurement collected furthest in time from the date of surgery) was evaluated. Other clinical and demographic characteristics in the 90 days before or day of surgery and resource utilization in the 90 days pre-surgery were analyzed. Descriptive statistics were reported as frequencies and means±standard deviations, and groups were compared using the Pearson chi-square test for categorical variables and Student’s t-test for continuous variables. Results:The total study population consisted of 2,984 TKR patients, of which, 9.1% (n=273) had an anemia diagnosis. In patients with an anemia diagnosis, the first anemia diagnosis in the EMR occurred 38±30 days before surgery. Patients with an anemia diagnosis versus those without an anemia diagnosis tended to be older (mean age 70.3±10.2 vs 67.5±10.0 years, respectively; p<0.001), have a lower proportion that were white (95.2% vs 98.5%, respectively; p<0.001), and have a higher comorbidity burden as measured by the Quan-Charlson Comorbidity Index (Q-CCI) (mean Q-CCI of 1.2±1.3 vs 0.6±1.0, respectively, p<0.001). The mean earliest Hb level for the total population was 13.7±1.3 g/dL and was <13 g/dL in 28.2% of TKR patients and <12 g/dL in 9.6% of TKR patients. The mean earliest Hb level was 12.5±1.3 g/dL in patients with an anemia diagnosis and was 13.8±1.2 g/dL for the group without an anemia diagnosis (p<0.001). Of patients with an anemia diagnosis, the majority (62.3%) had an earliest Hb level <13 g/dL (versus 24.8% of those without an anemia diagnosis; p<0.001) while 33.7% of those with an anemia diagnosis had an earliest Hb level <12 g/dL (versus 7.1% of those without an anemia diagnosis; p<0.001). Seven percent of patients with an anemia diagnosis had a hospitalization in the 90 days before TKR surgery versus 3.0% of those without an anemia diagnosis (p<0.001). In addition, 5.9% of the group with an anemia diagnosis had an ER visit versus 2.8% of those without an anemia diagnosis (p=0.005). In the 90 days pre-TKR surgery, the anemia diagnosis group had 5.0±2.7 days with an office/outpatient visit and 11.9±6.0 days with another-type service (e.g., prescription refill, administrative services) while the group without an anemia diagnosis had 3.6±2.0 (p<0.001) and 7.4±4.4 (p<0.001) days with these respective services. Conclusions:In this EMR database analysis, more than 9% of patients had an anemia diagnosis based upon ICD-9 codes in 90 days before or day of TKR surgery. Patients with an anemia diagnosis differed from those without an anemia diagnosis for certain characteristics including age, proportion who were white, comorbidity burden, and mean earliest Hb levels. The proportion of patients with a hospitalization in the 90 days preceding TKR surgery was significantly greater in patients with an anemia diagnosis as compared to those without an anemia diagnosis. In the total population, over 28% and 9% of patients had an earliest Hb value <13 g/dL and <12 g/dL, respectively, however, in patients without an anemia diagnosis, over 24% and 7% had an earliest Hb level <13 g/dL and <12 g/dL, respectively. This research provides further insight into preoperative Hb levels for a population of TKR patients with and without anemia diagnosis codes. Further research is warranted to better understand these variations between groups as well as the implications of these differences on preoperative management and post-operative outcomes in TKR populations. Disclosures:Forlenza:Janssen Scientific Affairs, LLC: Employment. Ellis:Janssen Scientific Affairs, LLC: Employment. Parise:Janssen Scientific Affairs, LLC: Consultancy. Lafeuille:Janssen Scientific Affairs, LLC: Consultancy. Lefebvre:Janssen Scientific Affairs, LLC: Consultancy.