Total hip arthroplasty (THA) is a highly successful orthopaedic procedure. Sir John Charnley, a British orthopedic surgeon known as the father of THA, developed the principles of hip joint replacement and designed a hip prosthesis in the 1960s (Erens & Thornhill, 2012). Total hip arthroplasty can eliminate joint pain, increase mobility, and improve a patient's quality of life (Erens & Thornhill, 2012). A wide variety of implants, as well as a variety of operative techniques to perform THA surgery, have been developed. There is also diversity in adjunct therapies associated with the THA procedure, two of which are drain use and blood management. This study examined THA drain and blood management practices at a Pacific Northwest hospital in the primary THA population to compare patient hemodynamic outcomes.

Literature Review

Drain Practices

Hippocrates is credited with the use of the first operative wound drain, a wooden tube, in 400 BC (Parker, Roberts, & Hays, 2004; Walmsley, Kelly, Hill, & Brenkel, 2005). Drains, commonly made of polyvinyl chloride (Cardinal Health, 2010) today, are a normal adjunctive therapy used in orthopaedic surgery. Because of the difficulty in achieving hemostasis in orthopaedic surgery, many orthopaedic surgeons utilize drains to prevent hematomas (Parker et al., 2004). Hematomas create wound tension, decrease tissue perfusion, and provide a hospitable culture environment for bacteria (Crevoisier, Reber, & Boesberger, 1998; Dora, von Campe, Mengiardi, Koch, & Vienne, 2007; Parker et al., 2004; Walmsley et al., 2005). Theoretically, drains prevent hematoma formation, thus preventing wound infection (a potentially devastating complication of joint replacement); however, drains may also act as bacterial conduits introducing infection into the surgical wound (Crevoisier et al., 1998; Dora et al., 2007; Parker et al., 2004; Walmsley et al., 2005). Parker et al. (2004) and Walmsley et al. (2005) further suggest that the drain material may inhibit a patient's natural immunity. In addition, drain use has been associated with increased wound drainage requiring patients with drains to receive more transfusions than patients who do not receive a drain after THA (Walmsley et al., 2005). The literature is rife with conflict regarding drain practice, but Crevoisier et al. (1998), Dora et al. (2007), Ritter, Keating, and Faris (1994), and Walmsley et al. (2005) concluded that drains provided no advantage and were not necessary in uncomplicated THA.

Drain practices at this Pacific Northwest hospital are oppositional. One group of surgeons routinely uses no postoperative drain on their THA patients and routinely orders a type and hold sample, or a type and cross match of two units of packed red blood cells (RBCs). The other group does not routinely order allogeneic blood, or obtain a type and hold or type and cross sample for primary THA patients but does routinely use the OrthoPAT. The OrthoPAT is an autotransfusion device that begins with the collection of intraoperative blood and continues with postoperative drainage collection via a drain tube from the operative wound connected to the OrthoPAT. The device collects, washes, and separates blood products into a concentrated hematocrit (HCT) product for reinfusion and waste products for discard (Haemonetics Corporation, 2003).

Autotransfusion Versus Banked Blood

There are a variety of autotransfusion devices on the market, some of which wash the collected blood and some that work on filtration principles alone (Warner, 2001). The autotransfusion system used at the study institution, the OrthoPAT, is initiated intraoperatively and continues for a maximum of 24 hours postoperatively (after a postoperative drain has been inserted intraoperatively and connected to the OrthoPAT device). Blood is collected in a reservoir and anticoagulated, washed with saline, and separated using centrifugal force. Concentrated RBCs are dispersed into an RBC bag and waste products are discarded into a waste bag. The American Association of Blood Blanks transfusion guidelines for all salvaged blood collected via the OrthoPAT are followed (Haemonetics Corporation, 2003), so a nursing intervention is required at least every 6 hours (in addition to the initial startup time investment, and end of surgery, end of recovery, and end of orthopaedic unit shift estimated blood loss [EBL] calculations) as the blood must be collected and transfused within a 6-hour window.

Cost-Effectiveness

In the spring of 2006, after an institutional joint program cost analysis determined that an autotransfusion protocol for THA surgery using the OrthoPAT was more cost-effective than an autologous predonation program, all THA patients (except those with infection, the presence of malignancy, or religious contraindications) were treated with autotransfusion (at the time of this decision the no drain group of physicians did not practice at this institution). Warner (2001) indicated that the OrthoPAT had been implemented at a community hospital for similar reasons, became the mainstay treatment for blood management after joint replacement surgery, and was considered a "safe, autologous option that requires no extra cost or time to the patient or institution" (p. 32). Clark, Spratt, Blondin, Craig, and Fink (2006) specifically recommend the perioperative use of the OrthoPAT to reduce the potential for complications from allogeneic transfusions in joint arthroplasty patients; however, they advocated for future research to analyze the potential cost savings to healthcare institutions as well as to clearly define which patients would likely benefit from OrthoPAT use.

Intraoperative blood salvage has been shown to be advantageous because it reduces the contaminants in shed blood (because it washes the blood). It also reduces the demand on blood bank resources as well as the need for banking blood. The product produced is primarily concentrated RBCs, and autotransfusion decreases the postoperative infection rate (Sinardi et al., 2005). Conversely, Lane and Crosby (2009) stipulated that at least two units of blood must be salvaged to make intraoperative cell savage cost-effective, making cell salvage a viable alternative in select procedures such as hip revisions, but not cost-effective for primary hip replacement due to the routine inconsequential blood loss in primary THA. Sinardi et al. (2005) agreed that the washing of salvaged blood is best used selectively when EBL is expected to exceed 600 ml. Tripkkovic, Sakic, Jakovina, Sakic, and Hrgovic (2010) concluded that utilizing a salvage device that did not wash the blood was safe and provided a good return product, though they offer that further studies need to be done to definitively recommend whether salvaged blood should be washed.

Rao, Dyga, Bartels, and Waters (2012) specifically evaluated transfusion costs related to the use of allogeneic RBCs, washed salvaged RBCs (the OrthoPAT was the preferable device used), and unwashed salvaged RBCs. The authors concluded that for primary elective THA, washed salvaged cells using the OrthoPAT (70.7% of THA patients received salvaged blood) was less costly than allogeneic transfusion but more costly than the transfusion of unwashed salvaged blood. Quality of blood collection is also a concern associated with autotransfusion because cell lysis may occur for a variety of reasons (Tripkkovic et al., 2010). Tripkkovic et al. (2010) evaluated the quality of blood collected in an autotransfusion device with low pressure suction (not the OrthoPAT) and determined that the RBCs sustained little damage. Rao et al. (2012) further determined that additional research was needed to clarify the medical necessity of a postoperative salvage system in primary elective THA since the majority of their THA patients received one unit or less of salvaged blood.

The OrthoPAT required a significant hardware and disposable investment by this Pacific Northwest hospital, as well as a significant, operationally essential, investment of nursing time (though an important concept, nursing time required to use the OrthoPAT was not evaluated over the course of this study). Education needs and time must also be assessed when implementing a device such as the OrthoPAT. Dora et al. (2007) recommended that cost-reduction efforts must address the nursing time required to maintain cost-effective therapies.

Allogeneic Blood Transfusions

In a recent study, red cell transfusion rates were evaluated in 1,827 orthopaedic surgical procedures (1,787 patients undergoing several different orthopaedic procedures) and 53% of the THA patients received a transfusion (Verlicchi, Desalvo, Zanotti, Morotti, & Tomasini, 2011). Rao et al. (2012) reported that transfusion rates in elective total hip and knee arthroplasty surgery can vary from 22% to 97%, and according to Widman, Jacobsson, Larsson, and Isacson (2002), hip replacement patients with drains required more blood transfusions.

Transfusion of allogeneic blood is a standard therapy to correct THA operative blood loss (Kleinert, Theusinger, Nuernberg, & Werner, 2010; Lane & Crosby, 2009), yet it is not without its disadvantages. The disadvantages of RBC transfusion include: the scarcity and finite limits of this precious resource; patients prefer not to be transfused in spite of the increased sensitivity and sophistication of screening procedures applied to the blood supply; and blood transfusion is associated with a postoperative increase in length of stay and morbidity (Lane & Crosby, 2009). In addition, the immunomodular effects of blood transfusions are thought to increase the occurrence of postoperative infections (Weber et al., 2005).

The cost of allogeneic transfusion is also a consideration. Kleinert et al. (2010) indicated that a unit of packed red cells ranged in price from $270 to $780 (price variability was attributed to laboratory tests that may be done on each unit and storage fees). A more recent report priced a unit of allogeneic RBCs between $522 and $1,183 (Rao et al., 2012). The total cost of transfusing a unit of red cells at this Pacific Northwest study institution (price and total cost include the red cell unit price plus processing and blood administration fees) exceeds $1,000. All these findings support the necessity to avoid routine RBC transfusion and to transfuse THA patients only when clinically indicated (Guerrin, Collins, Kapoor, McClean & Collins, 2007). Guerrin et al. (2007) stated "that the challenge lies in selecting the best blood management strategies for each patient" (p. 40).

Problem and Purpose

This study was inspired by nursing observations. The principal investigator (PI) and other orthopaedic nurses questioned the differences in drain practices between surgeon groups, as well as patient outcomes in the different THA groups. The purpose of this study was to compare hemodynamic outcomes of two patient groups: (1) primary THA patients with a postoperative drain connected to the OrthoPAT, an autotransfusion device, and (2) primary THA patients without a postoperative drain. The three objectives of this study were as follows:

* To evaluate the use of banked blood in primary THA patients;

* To analyze the volume of return and rate of wastage of autologus blood collected using the OrthoPAT; and

* To compare hemodynamic status, specifically HCT percentages, preoperatively and postoperatively in the identified two groups of primary THA patients.

The intent of the study was to gather data to inform the surgical team and hospital management about the usage and cost of autotransfusion devices, transfusion practices and the cost of allogeneic blood, and THA patient outcomes.

Materials and Methods

This descriptive quantitative study was conducted by means of a retrospective medical record review at an acute care hospital in the Pacific Northwest. Internal review board approval was obtained first from the PI's educational institution and second from the Pacific Northwest hospital's internal review board. This retrospective chart review involved no direct patient contact and minimal risk. A convenience sample of THA patient medical records was obtained from more than 500 THAs annually performed at this hospital. Inclusion criteria for the study were:

* Patients who had THA on the identified operative hip for the first time.

* Patients recognized in the medical record with the same admitting diagnosis.

* Patients who received an uncemented prosthesis.

* Patients who received the routine hip implants typically used by their surgeon via the surgeon's standard technique.

* Patients who received a drain connected to the OrthoPAT or who did not receive a drain.

Fractured hip patients and revision hip patients were excluded from this study because of the dissimilar nature of their procedures, varying admitting diagnoses, and the implant, duration, and EBL variability of hip revision surgery. The patient medical records of the three orthopaedic surgeons with the highest primary total hip volume for both Groups A and B were chosen for this study for a total of six surgeons (due to the disparity of case volume between the surgeon groups a total of 80 [N = 80] records was the originally chosen quota to maintain equity in the total number of patient charts reviewed). Within Groups A and B, 14 patient medical records were selected from the one surgeon with the highest THA volume and 13 patient medical records from each of the remaining two surgeons, for a total of 40 medical records from each group. Every third chart was selected until the maxim number of charts available was obtained or the quota for the specific surgeon had been reached. Each chart was assigned an identification number to ensure that HIPPA confidentiality was maintained, then each chart was reviewed.

Multiple demographic variables such as gender, laterality, age, height, weight, and body mass index (BMI) were assessed along with the admitting diagnosis. Intraoperative data captured the procedure time, anesthesia score and method, and temperature across the perioperative course. Hemodynamic data gathering was centered on serial HCT percentages and prothrombin and international ratio coagulation factors. Postoperative chemical anticoagulation practice was also evaluated. Two data sections were designed to capture transfusion practices and the transfusion section completed was dependent on surgeon drain practice. Comorbidity commonalities were also assessed across the THA population. Data were evaluated to examine relationships between variables, evaluate transfusion practices, and analyze outcomes between those THA patients receiving a drain with an autotransfusion device versus those who did not receive a drain.

Statistical Analysis

The data analysis was completed using Statistical Package for the Social Sciences (SPSS) version 18.0 (SPSS Inc., Chicago, IL). Levene's test, two-tailed t test, and analysis of variance (ANOVA) were conducted.

Results

There were a total of 74 (N = 74) patient records selected for study inclusion, 40 (n = 40; 15 male and 25 female) from Group A (drain with autotransfusion device) and 34 (n = 34; 16 male and 18 female) from Group B (no drain). Patients were compared in five main categories: (1) BMI, (2) American Society of Anesthesia (ASA) score, (3) comorbidities, (4) intraoperative fluid volumes infused and EBL, and (5) HCT.

Demographic Comparisons

The mean age of Group A patients was 64.25 (median = 63.50, standard deviation [SD] = 9.254, minimum 43, and maximum 82) years, and the mean age of Group B patients was 65.26 (median = 63.50, SD = 10.983, minimum 45, and maximum 88) years. The BMI distribution between groups varied slightly with the majority of Group A patients (47.5%) classed as obese whereas the majority of Group B patients (38.2%) were classed as overweight (see Table 1). When ASA scores were compared, both Group A and Group B were similarly classed (see Table 2). The primary comorbidities included Diabetes Type II, hypertension, hypercholesterolemia, and obesity. The sample demonstrated similar rates of occurrence of Diabetes Type II and hypertension, but the groups differed slightly in the incidence of hypercholesterolemia and obesity (see Table 3). The intraoperative fluid volume infused (see Table 4) for both groups was similar based on the mean volume; however, the range for fluid volume infused in Group B was more widely dispersed than that in Group A. The EBL for both groups was nearly equal (see Table 4; the EBL for Group A is listed as a combined EBL because the intraoperative and postanesthesia recovery unit EBLs were added together to get the combined OrthoPAT EBL for the perioperative period). Demographic variable analysis demonstrated no statistically significant difference between groups.

Table 1. CDC BMI Classification Distribution

Table 2. ASA Distribution

Table 3. Comorbidity Distribution

Table 4. Intraoperative Fluid Volume and EBL

Transfusion Results

In Group A, patients receiving a drain connected to the OrthoPAT autotransfusion device, 27 of 40 patients (67.5%) received an OrthoPAT collected transfusion. The transfusion volumes noted ranged from a low of 50 cc to a high of 425 cc, and transfusion volumes of 50 cc (five cases) and 100 cc (five cases) were the most frequently reported volumes. Five patient operating room (OR) records had a recorded transfusion volume of 0, and eight of 40 patient OR records (20%) did not have a transfusion volume recorded. One patient in Group A was typed and crossed for two units of packed cells and received a one-unit transfusion with a noted HCT at the time of transfusion of 26.9%. Eighteen of 40 patient OR records noted OrthoPAT waste volumes that ranged from a low of 10 cc to a high of 130 cc. The remaining 26 records did not have an OrthoPAT waste volume recorded.

In Group B, patients with no drain, one patient was typed and screened preoperatively, 28 of 34 (82.4%) patients were typed and crossed for two units of packed cells, and five patients were neither typed and screened nor typed and crossed preoperatively. Eleven of 34 patients (32.4%) were subsequently transfused with pretransfusion HCTs ranging from a low of 24% to a high of 28.9%. Group B patients received a total of 24 units of packed cells (transfused patients primarily received two units; however, one patient did receive a total of 6 units).

HCT Analysis

Table 5 summarizes the HCT results for three time frames: preoperative/day of surgery, postoperative day (POD) 1, and POD 2. The distribution of HCT percentages for all three time frames is similar between groups. Levene's test for equality of variance demonstrated equality of variances between Groups A and B. A two-tailed t test (see Table 6) was used to evaluate differences between groups for HCT status preoperative/day of surgery and on POD 2, and the results led the PI to retain the null hypothesis: there is no difference in HCT status preoperatively or postoperatively between groups. An analysis of variance level (see Table 7) did not demonstrate significant differences in any of the three HCT measurements between groups.

Table 5. Hematocrit

Table 6. Two-Tailed

Table 7. Analysis of HCT Variance

Discussion

Forty eligible patients were obtained for Group A. Thirty-four qualifying records were selected for inclusion for Group B. The quota of 40 patient records was not met for Group B for two reasons: (1) the reduced case volume in Group B compared with the case volume in Group A and (2) because there was a higher rate of drain use than expected. Surgery demographics (age, BMI, ASA, and comorbidities) were similarly represented in both groups. In addition, reported intraoperative fluid volumes and EBLs were comparable between groups essentially establishing equality of the two groups across multiple factors both before and during THA.

OrthoPAT

The rate of return on the OrthoPAT blood was higher than expected, considering that the general nursing impression was that very little autologus blood was being transfused. There was also a wide variability in the amount of OrthoPAT blood returned among Group A patients; however, the most frequently recorded volumes over the 24-hour postoperative period during which the autotransfusion device was operated were low. Retrieval of the OrthoPAT data points was difficult at times because of the inconsistency in charting and the lack of legibility of some data entries. Because only volumes that were clearly noted in the nursing notes were used in the analysis, volume data were not obtained from all 40 patient records in Group A, reducing the amount of data available to be analyzed.

Perioperative blood management using intraoperative blood salvage devices like the OrthoPAT is a recognized and supported component of joint replacement surgery, and its use enabled the elimination of a autologus pre-donation program at this Pacific Northwest Hospital. While the benefit of the therapy was evaluated prior to its introduction into practice, the results suggest the need for a current cost-benefit analysis. Furthermore, because there was no significant difference in HCT outcomes between the THA groups, the PI would recommend an examination of OrthoPAT practice to explore potential practice changes.

HCT Status

The preoperative HCTs of both groups were similarly distributed and the postoperative HCT results of the groups were not significantly different. The no significant difference finding was, in fact, significant because the autotransfusion device is used on the basis of the understanding that the overall blood loss is less than it would be without the autotransfusion device (because blood is collected intraoperatively, washed, and returned) and the fact that the blood product returned via the washing process is of a high HCT. This should, in turn, reduce the postoperative HCT drop, but Group A did not have higher postoperative HCT results than Group B. This suggests that using the OrthoPAT on all primary THA patients may not be a cost-effective blood management treatment because of (a) the significant institutional hardware and disposable investment, (b) the required nursing time to run the device, and (c) the lack of significant patient benefit.

Transfusion Practices

On the basis of data analysis, Group B received more allogeneic units (32.4% transfusion rate; the transfusion rationale was not evaluated) than Group A; however, 67.5% of Group A patients received an OrthoPAT transfusion without yielding a significant difference in HCT results between the two groups. The study results suggest that there may be variability in allogeneic transfusion practice generating the need to define current transfusion parameters. Once the current state is defined, the joint team can deliberate how to improve and standardize the transfusion practice to generate a reduction in the overall transfusion rates.

Nursing Implications

A study like this one, which was inspired by nursing observations, can generate evidence to support practice change to enhance patient outcomes. Nurses working collaboratively with the total joint team would need to evaluate the study results and then explore alternative therapies such as the preoperative administration of tranexamic acid to reduce total joint replacement surgery intraoperative blood loss and postoperative allogeneic transfusions (Rajesparan, Ahmad, & Field, 2009; Snyder et al., 2012); and the importance of detecting preoperative anemia early so that it can be treated adequately prior to surgery to reduce the transfusion requirements of anemic patients (Kleinert et al., 2010). A thorough evaluation of current practice as well as the benefits of alternative therapies could lead to improvements in the judicious use of resources, and patient outcomes.

Future Research

A prospective study with a larger sample size would be beneficial in evaluating autotransfusion devices and the no drain practice to analyze the postoperative HCT status of all THA patients. A study of this nature may better define best practice for autotransfusion drain use and intraoperative and postoperative management of blood loss to minimize the usage of allogeneic blood resources. In addition, further investigation into alternative therapies for blood management in THA should be evaluated to determine their potential benefits across the THA population to establish a cost-effective and beneficial blood management and treatment protocol.

Limitations

A power calculation was not performed to determine sample size; instead, a small convenience sample was used to conduct this study. Patient medical records were randomly chosen from select physicians. Not all physicians performing THA at the chosen hospital site were included in this study, so study results may not be representative across all primary THA populations. In addition, the quota was not met for Group B because of the limited number of cases conducted in the identified selection period and the number of cases that qualified for selection within the cases completed. Study generalizability may be limited because the study sample was obtained from one acute care hospital.

Conclusion

Osteoarthritis is a common, painful, and sometimes debilitating condition that will drive more and more patients to seek THA when conservative treatment fails. Divergent drain and blood salvage practices practiced by physicians performing THA will continue until well-designed prospective studies demonstrate significant benefits of one therapy versus the other. In addition, further investigation into alternative therapies for blood management in THA should be evaluated to determine their potential benefits across the THA population. The best drain and blood management therapies remain undefined, but it is clear that chosen therapies within the total joint program need to prevent the overuse of allogeneic resources, be cost-effective, and provide significant patient benefits.

Acknowledgments

The author thanks Dr. K. Fitzsimmons and Dr. S. Casey for their guidance during the research process.

References