One of the potential negative side effects of joint replacement surgery (JRS) is site infection (SI). While a well-evidenced approach to the issue is the use of antibiotics, there is little consensus on taking into account the impact of body mass index (BMI) while determining the dosage (Cinotti et al., 2018). This proposal will present a retrospective observational study (chart review) that will investigate the correlations between different doses of antibiotics, BMI, and JRS-related SI.
The paper will include a short overview of the literature that addresses the topic, a discussion of the study’s framework, and a statement of research hypotheses. The methodology, including plans for data collection and analysis, will be considered next, and an approximate budget and timeline will be introduced. Finally, a conclusion will report the specifics of the proposed project. The goal of this proposal is to request funding from the Boston Medical Center (BMC).
JRS is an important and very common operation. In 2011, over 1.2 million joint replacement procedures took place in the US; among them, total knee replacements were the most numerous operations, and hip replacements followed suit (Berríos-Torres et al., 2018). The likelihood of SI development as a result of JRS may be estimated at about 1-2.18%; they are not the most common outcome of the procedure (Berríos-Torres et al., 2018; Marculescu, Mabry, & Berbari, 2016).
However, this number has been growing, and it is projected to achieve 6.5-6.8% by 2030 (Berríos-Torres et al., 2018). In addition, SI is a very unfavorable development that carries significant risks for patients, prolongs hospital stays, and increases expenses (Marculescu et al., 2016). Overall, SI is a major complication, and it is important to prevent its development.
The prevention of infections involves a complex of activities (from risk assessment to hand hygiene to surgical technique). However, a primary solution that is evidenced to be very effective is the use of antibiotics, including their intravenous infusion before the surgery; this approach is termed as antibiotics prophylaxis (Marculescu et al., 2016). The application of antibiotics requires the consideration of benefits and risks, as well as harms (Branch-Elliman et al., 2017).
Furthermore, the effectiveness of antibiotics is determined by their timely application, as well as the adjustment of the dose to a patient’s BMI (Wu et al., 2016). Put simply, insufficient amounts of antibiotics are unlikely to be effective, but their excessive use is associated with risks (Sanders, Goslings, Mathôt, & Schepers, 2019). Currently, the appropriate antibiotics prophylaxis is being actively researched since it differs for various types of surgeries (Berríos-Torres et al., 2018). However, the question of the interaction of BMI and doses of antibiotics is still being investigated (Cinotti et al., 2018). Therefore, the exists a gap in the current knowledge on a very important complication of a rather common surgical procedure, which calls for its elimination.
It is also noteworthy that multiple factors increase the likelihood of SI after JRS. They include the specifics of surgery (for example, being a repeat surgery or taking more time than expected) and certain patient factors (including the patients’ health, for instance, nutritional state or glycemic status) (Buetti et al., 2018; Marculescu et al., 2016).
The latter group incorporates BMI (Meijs et al., 2019; Wilson et al., 2018); patients with the BMIs that are classified as excessive weight and obesity are considered to be at a greater risk of infections. It should be pointed out that the possibility of them being underdosed when it comes to prophylaxis is an issue (Wilson et al., 2018). In other words, it is possible that the correlation between BMI and SI is at least partially explained by insufficient antibiotics dosages. However, this correlation is still noteworthy because it highlights the importance of investigating SI and antibiotic prophylaxis in patients with larger BMIs.
Problem and Significance
The presented evidence spells the problem out in the following way. Patients with larger BMIs are more likely to develop SI, and they are not unlikely to receive insufficient doses of antibiotics for their JRS (Wilson et al., 2018), but the question of BMI-appropriate doses is still being researched (Cinotti et al., 2018). Given the increasing importance of JRS and increasing likelihood of JRS-related SI (Berríos-Torres et al., 2018), the additional investigation of the topic is justified and significant.
The Data-Information-Knowledge-Wisdom (DIKW) framework was initially introduced as a nursing informatics theory. These days, it can be applied to nursing and healthcare as such, especially since information systems in healthcare are becoming ubiquitous (Ronquillo, Currie, & Rodney 2016). For this project, the framework’s focus on the process of the transformation of raw data into applied knowledge (wisdom) is important. The data about SI infections in people with different BMIs and after surgeries with different dosages of antibiotics exist within the many charts of patients who underwent JRS.
This project will accumulate and organize the data and proceed to interpret it, upgrading it to the level of information about SSI and its relationships with the two other variables. Future studies will be able to integrate this information further (upgrading it to knowledge) and apply it (upgrading it to wisdom). These are the key terms of the framework, as described by Ronquillo et al. (2016), and they fit the processes of the proposed research.
Purpose, Research Hypotheses, and Questions
The purpose of the project is to attempt to determine JRS-related SI incidence variations that are associated with the doses of antibiotics and BMI. The following question follows: are there any variations in SSI incidence that are correlated with the dosages of antibiotics and patients’ BMIs? The first independent variable is the dosage of antibiotics; its introduction is what is likely to affect (or fail to affect) the dependent variable, which is SI. The second independent variable is BMI; it is also expected to be associated with SI or its absence. Thus, there are three null hypotheses.
- There is no correlation between antibiotics dosages and SI incidence in patients undergoing JRS.
- There is no correlation between BMI and SI incidence in patients undergoing JRS
- There is no interaction between BMI and antibiotics dosages in their effect on SI incidence in patients undergoing JRS.
Conversely, the hypotheses would consist of claims that antibiotics (different dosages) and BMIs (different BMIs) are going to show a correlation with SIs; specifically, higher BMI would be associated with more infections, and higher dosages in people with higher BMI would result in fewer infections. These hypotheses are supported by the presented literature, especially by Wu et al. (2016).
The planned methodology consists of the following elements and characteristics. First, the study has to be quantitative since it looks for correlations between variables, and only quantitative research can perform this task (Gray, Grove, & Sutherland, 2016; Polit & Beck, 2017). Second, the presented project will be a correlational (observational) and retrospective study (Sarkar, 2014). In other words, this research will not presuppose any interaction with any participants, and no direct manipulation of a variable will take place (Bauman, Jackevicius, Zillich, Parker, & Phillips, 2018; Polit & Beck, 2017).
Instead, the study will observe the conditions and outcomes in the patients who have already had their joints replaced. The primary reason for this choice is that there are significant procedural and ethical concerns attached to experimentation on people who are in need of surgery. From this perspective, a non-experimental study is more ethical and simpler to arrange.
To clarify, the project will be a retrospective chart review. It will employ already existing charts to collect data about JRS, including antibiotics, BMI, and presence or absence of SI. The primary investigator will get access to the Orthopedic Surgical Center (OSC) of BMC. The decision is very defensible; first, this approach can provide all the necessary information for observing the three variables in the studied population.
Therefore, it can be used to provide some data that will contribute to responding to the research question. Second, this approach allows accessing a rather large body of data. The alternative to retrospective studies is prospective studies, but for chart reviews, prospective studies have access to fewer charts that are limited by the time of observation (Gray et al., 2016). A retrospective review has access to many more charts and does not have to involve any waiting time; the proposed study will review five years’ worth of charts, but it will only last six months.
Finally, retrospective chart reviews do not involve recruitment procedures for patients and involve zero risks for their health, which implies that they are easier to keep ethical than intervention-based studies. In fact, it is not uncommon for such studies to be exempt from a review by an Institutional Review Board or receive an expedited review rather than a full one (Bauman et al., 2018). The confidentiality-related risks remain a concern, however, and it will be necessary to ensure the appropriate de-identification of any charts involved (Bauman et al., 2018; Sarkar, 2014). The relevant procedures will be accounted for before the project is considered by a Review Board.
However, it should be acknowledged that significant challenges are associated with non-experimental and retrospective studies. First, non-experimental studies are not the same level of evidence as experiments or quasi-experiments (Polit & Beck, 2017). Second, they are noninterventional, which limits their ability to suggest cause-and-effect relationships between variables (Gray et al., 2016). As a result, this project explicitly focuses on correlations, which means that this limitation is reflected in the research question, hypotheses, and design. Second, retrospective studies are typically subject to a greater possibility of error than prospective ones (Houser, 2016; Polit & Beck, 2017).
Indeed, the data were not recorded specifically for the project, and no researcher-instituted precautions or controls for their recording were present (Gray et al., 2016; Polit & Beck, 2017). The proposed research will use charts, which are likely to be accurate, and the data will be cleaned to check for impossible information as a form of precaution (Gray et al., 2016). Still, this limitation will need to be considered in the review of the findings.
The proposed sampling method is probability sampling because probability samples lead to better, more generalizable findings. This feature is important for a correlational study (Gray et al., 2016; Polit & Beck, 2017). Furthermore, for this project, BMI variances are important for the sample, which is why it is necessary to ensure their representation in the sample. Therefore, stratified probability sampling is going to be used (Houser, 2016; Jackson, 2016). Data collection in retrospective chart reviews is conducted with the help of a protocol, which is developed to include a project’s variables. As for data analysis, when multiple variables are considered, multivariate analysis is usually required (Polit & Beck, 2017). The plans for sampling, data collection, and data analysis will be covered extensively in the next section.
Plans for Data Collection and Analysis
Approximate plans for the key procedures of the project have been prepared. A retrospective analysis of charts is the specific method of data collection that will be employed. It is planned to request de-identified charts from BMC; this way, the research will avoid most of the ethical issues of a chart review. The chart review itself will be performed by the primary investigator, although it is also planned to recruit an assistant reviewer from the BMC medical team.
Stratified random sampling will be applied to the charts of the customers of the OSC of BMC. Currently, it is planned to review the charts of the patients for the past five years, and the desired goal is to select 30 patients for each year. Thus, the total sample will consist of 120 patients. The stratified categories will include people with normal BMI, people who have a BMI that is classified as overweight, and people with different classes of obesity. Regarding selection criteria, the studied population consists of people who underwent total JRS. Patients will be eligible regardless of their gender and age, but as a major confounding variable, repeat surgeries will be excluded. Repeat surgeries are a risk factor for SI (Buetti et al., 2018; Marculescu et al., 2016), which is why the removal of this variable is preferable.
Thus, the sampling process can be described as follows. First, the charts of the people who had first-time total JRS will be found. They will be de-identified by the hospital; most likely, the recruited assistant investigator will perform the task. After de-identification, they will be available to the researcher who will carry out the rest of the sampling procedures. The charts will be categorized based on their BMIs.
Equal numbers of people from different BMI categories will be randomly selected with the final goal of selecting 30 charts per year for the past five years. The project will use a data extraction form that will be created based on the studied variables and specifics of the site’s charts. It is planned to make it an electronic form since they facilitate data extraction (Sarkar, 2014). This form will serve as the project’s data collection instrument, and it will be used to extract the data pertinent to the presence or absence of SI, the patient’s BMI, and the surgery antibiotics dose.
As it was mentioned, the appropriate data analysis approach for the described methodology would be multivariate analysis. ANOVA and ACOVA are the most common choice, but they are best used with interval or continuous data (Gray et al., 2016; Polit & Beck, 2017). BMI and dosages can be considered continuous, but infection incidence uses a nominal scale (yes/no). Therefore, logistic regression is the best choice because this method of analysis works with binary and continuous variables (Polit & Beck, 2017). The chosen approach will help to determine the probability of SI in the groups with different BMI and dosages, and it will also identify interactions between variables. In general, the proposed methods were selected to ensure their ability to test the hypotheses and fit into the design of the project.
Budget and Timetable: Feasibility
Chart reviews are not very costly; the only crucial expenses for them are the pay for the investigators’ work. For this research project, only one independent investigator will be recruited, and they will be paid their usual hourly wage. In addition, a research coordinator will be involved, and they will use the same pay rate. Finally, the principal investigator’s wages are also going to be included in the budget. Based on the current information about the remuneration practices of BMC, the usual hourly wage for the nursing staff amounts to about $43. The project is unlikely to take up full shifts; rather, it will be a part-time job with up to 20 hours of work per week.
While it is unlikely that all the participants will be engaged for six months, it is still proposed to calculate the wages for the entirety of the project to determine the maximum possible expenses. As a result, the maximum anticipated wage for one nurse would be $20,640, and to calculate the anticipated maximum budget of the project, the figure needs to be tripled to account for the three specialists involved. Thus, it is expected to spend no more than $61,920 on this project’s activities (see Table 1).
Table 1. Maximum Anticipated Budget for the Project.
|Maximum Hours/Week||Total Maximum Hours||Wage/Hour ($)||Final Wage per Researcher||Number of Researchers||Total Maximum Expenses ($)|
As for the timeline, the most time-consuming processes of this chart review will be data collection and analysis. Sampling will also take some time due to the decision to employ the stratified random approach. In addition, the review of relevant literature will be carried out as needed, although it will be especially important while preparing the proposal and the final report. As a result, for a six-month timeframe that is planned, the approximate timetable in Figure 1 is proposed. To summarize, this section demonstrates that the proposed research is feasible and can be conducted within a limited timeframe with only three specialists.
With JRS becoming more widespread and more SI-prone, the investigation of its understudied aspects becomes more important. In particular, it is necessary to determine the correlations between different dosages of antibiotics, patient BMIs, and JRS-related SIs. Given that BMI is a risk factor for JRS-related SIs and that people with greater BMIs are frequently underdosed with antibiotics, the significance of the topic is apparent. The presented project will be a quantitative retrospective chart review that will use stratified probability sampling and a self-developed protocol for data extraction. The DIKW framework will be used to direct the procedures.
Appropriate statistical methods will be applied as well, which means that logistic regression will be used due to the specifics of the variables. There are limitations to the methodology, but it can be used to reject or support the hypotheses and incorporates ethics-related safeguards, especially those related to chart de-identification. Finally, the proposed plan is feasible, does not require too many specialists or resources, and can be carried out within six months. Given the importance of the presented research problem, its investigation with these methods and budget is fully justified.
Bauman, J., Jackevicius, C., Zillich, A., Parker, R., & Phillips, B. (2018). On the methodology of retrospective chart reviews. Journal of the American College of Clinical Pharmacy, 2(1), 6-7. Web.
Berríos-Torres, S. I., Umscheid, C. A., Bratzler, D. W., Leas, B., Stone, E. C., Kelz, R. R.,… Dellinger, E. P. (2017) Centers for Disease Control and Prevention Guideline for the prevention of surgical site infection, 2017. JAMA Surgery, 152(8), 784-791. Web.
Branch-Elliman, W., Ripollone, J., O’Brien, W., Itani, K., Schweizer, M., & Perencevich, E. et al. (2017). Risk of surgical site infection, acute kidney injury, and Clostridium difficile infection following antibiotic prophylaxis with vancomycin plus a beta-lactam versus either drug alone: A national propensity-score-adjusted retrospective cohort study. PLOS Medicine, 14(7), e1002340. Web.
Buetti, N., Atkinson, A., Troillet, N., Eisenring, M. C., Zwahlen, M., Kuster, S. P.,… Marschall, J. (2018). 2137. Risk factors for surgical site infection after joint replacement surgery: Data from the Swiss National Surveillance System. Open Forum Infectious Diseases, 5(suppl_1), S629-S629. Web.
Cinotti, R., Dumont, R., Ronchi, L., Roquilly, A., Atthar, V., Grégoire, M.,… Asehnoune, K. (2018). Cefazolin tissue concentrations with a prophylactic dose administered before sleeve gastrectomy in obese patients: A single centre study in 116 patients. British Journal of Anaesthesia, 120(6), 1202-1208. Web.
Gray, J. R., Grove, S. K., & Sutherland, S. (2016). Burns and Grove’s The practice of nursing research-e-book: Appraisal, synthesis, and generation of evidence (8th ed.). Philadelphia, PA: Elsevier Health Sciences.
Houser, J. (2016). Nursing research: Reading, using and creating evidence (4th ed.). Burlington, MA: Jones & Bartlett Learning.
Jackson, S. L. (2016). Research methods and statistics: A critical thinking approach. (5th ed.). Belmont, CA: Wadsworth Cengage Learning.
Marculescu, C., Mabry, T., & Berbari, E. (2016). Prevention of surgical site infections in joint replacement surgery. Surgical Infections, 17(2), 152-157. Web.
Meijs, A., Koek, M., Vos, M., Geerlings, S., Vogely, H., & Greeff, S. (2019). The effect of body mass index on the risk of surgical site infection. Infection Control & Hospital Epidemiology, 40(9), 991-996. Web.
Polit, D.F., & Beck, C.T. (2017). Nursing research: Generating and assessing evidence for nursing practice (10th ed.). Philadelphia, PA: Lippincott, Williams & Wilkins.
Ronquillo, C., Currie, L., & Rodney, P. (2016). The evolution of Data-Information-Knowledge-Wisdom in nursing informatics. Advances in Nursing Science, 39(1), E1-E18. Web.
Sanders, F., Goslings, J., Mathôt, R., & Schepers, T. (2019). Target site antibiotic concentrations in orthopedic/trauma extremity surgery: Is prophylactic cefazolin adequately dosed? A systematic review and meta-analysis. Acta Orthopaedica, 90(2), 97-104. Web.
Sarkar, S. (2014). Conducting record review studies in clinical practice. Journal of Clinical and Diagnostic Research, 8(9), 1-4. Web.
Wilson, C. J., Georgiou, K. R., Oburu, E., Theodoulou, A., Deakin, A. H., & Krishnan, J. (2018). Surgical site infection in overweight and obese total knee arthroplasty patients. Journal of Orthopaedics, 15(2), 328-332. Web.
Wu, C., Chen, I., Wang, J., Ko, J., Wang, C., & Lee, C. (2016). Surgical site infection after total knee arthroplasty: Risk factors in patients with timely administration of systemic prophylactic antibiotics. The Journal of Arthroplasty, 31(7), 1568-1573. Web.