Additionally, patients who may be traumatically injured and who are not admitted to a participating trauma center will not be included in the NTDB, nor will trauma patients who died on scene before being transported. Another consideration to note is that participating hospitals may differ in their criteria of which patients to include in the database, specifically patients who are dead on arrival or those who die in the Emergency Department . This discrepancy in inclusion and exclusion criteria between hospitals regarding specific injuries makes representative comparisons potentially difficult. Lastly, it is important to mention that large databases such as the NTDB are subject to missing data or disparate data. This is often due a multitude of factors, a few of which various demographic data points, test results and other key information, such as procedures, that may not be documented in the health record and therefore omitted in the database . Missing data often contributes to information bias; however, it can also contribute to selection bias because one of the methods in dealing with missing data is excluding participants for which data is missing thereby creating potential selection bias. Missing data may undermine the ability to make valid inferences, therefore, steps will be taken throughout the design and operational stages and methods within this study to avoid or minimize missing data. Methods to reduce information bias that can lead to selection bias will be discussed in the analysis section of this paper.Due to the methods by which data are collected and inputted into the NTDB,ebb and flow flood table potential problems are encountered in terms of data accuracy. Under reporting of variables obtained from the NTDB has often been noted as a problem due to the reliability of data extraction by participating hospitals .A major variance between participating hospitals is that hospitals with more resources are more likely to have dedicated staff to data collection.
This can lead to informational bias in those hospitals that are more compliant in reporting data metrics when compared to others that are not. For example, hospital data registries that have incomplete data on complications may appear to deliver better care than hospitals that consistently record all complications. A recent study by Arabian et al. revealed the presence of inaccuracy and variability between hospitals, specifically in the areas of data coding and injury severity scoring. Additionally, the type of registry software a hospital utilizes can report injury severity scores differently . This too, renders data subject to informational bias. Information bias is due to inaccurate or incorrect recording of individual data points. When continuous variables are involved, it is called measurement error; when categorical variables are involved, it is called mis-classification . In this study, the potential for information bias is mostly due to 1) incomplete data documented in the medical record, or 2) inaccurate entry into the hospital trauma database by hospital staff. Missing data will be analyzed in terms of potential effect for both the independent and dependent variable . While the database captures marijuana exposure through the first recorded positive drug screen within the first 24 hours after first hospital encounter, it is recognized that at times patients will not be screened, even if they have been exposed to marijuana. Marijuana exposure is identified through the presence of Cannabinoid in a urine toxicology screen. Marijuana presence can be detected in the urine up to 3-5 days from exposure in infrequent users; marijuana can be detected up to 30 days for chronic users . Therefore, patients could potentially have a positive marijuana toxicology screen even though they may not have ingested marijuana the day of the event. A positive marijuana urine toxicology screen indicates the probability of prior use, not immediate use. This is an important limitation to note. In clinical practice, the determination for a toxicology screen is often symptomology, so it is reasonable to assume that patients who have ingested marijuana a week prior to the event date may not exhibit the expected symptomology. Unlike other observational cohort studies, the potential of recall bias is minimal due to the availability of an objective marker to measure the independent variable, namely, the presence of marijuana.
The presence of marijuana is captured from the hospital lab urinalysis results and is recorded as present within 24 hours after the first hospital encounter. Similarly, the data entered to measure the GCS score is also captured objectively through a numeric recorded score found in the medical record. See analysis section for how this type of bias will be addressed. The Trauma Quality Programs research database housed in the NTDB for the year 2107 is the time frame for this study. Though initially the researcher intended to include data from 2013-2017, data from years other than 2017 had to be excluded. In effort to standardize the type of data collected by local, regional, and state trauma registries, the NTDB designs a National Trauma Data Standard Data Dictionary that is designed to establish a national standard for the collection of trauma registry data while also providing the operational definitions for the NTDB. In summary, the NTDS provides the exact standards for trauma registry data submitted to the NTDB. Prior to the 2017 data dictionary, trauma registry programs had limited selections regarding data related to drug use. The options provided by the NTDB registry only included whether drug use was present and whether it was confirmed by a test or by prescription. It did not allow the trauma data abstractor to specifically identify the type of drug found. In 2017, the data dictionary was revised to include a drug screening category that aimed at recording the first positive drug screen result within 24 hours after the first hospital encounter. Typically, in trauma hospitals reporting to NTDB and within the context of trauma, acquisition of a urine and blood drug and alcohol screen is standard expectation of practice. It then provided a list of 15 options for the abstractor to choose from. Because it was impossible to isolate cannabinoid use in earlier data sets, the researcher was only able to use the 2017 NTDB data set, which at the beginning of the study was the latest available data set by the NTDB. As of February 13th, 2021 the 2018 NTDB data set was not available. All the trauma data used in this study are organized by an element INC_KEY, which is a designated unique identifier for each record. The designated unique identifier INC_KEY expresses a unique clinical visit/episode by an individual at a participating trauma center. It is important to consider that an individual could have been included/counted more than once in the registry because of more than one traumatic event within the year.
The Participant Use File Trauma data set contained all the demographic, environmental, and clinical data information. However, it did not identify or delineate TBI cases as such. Therefore, a separate data set that contained ICD 10 Diagnosis Codes had to be utilized to identify TBI cases which then could be used to create a merged data set that is complete. The 2017 PUF Trauma data set was uploaded to SPSS version 25 on September 10th, 2020. The PUF Trauma data set included a total of 997,970 unique identifier cases. A frequency analysis was performed to ensure no duplicate cases were found . The PUF Trauma data set included 328 unique variables. Next, the PUF ICD-10 Diagnosis data set was uploaded and examined. The PUF ICD Diagnosis data set is organized via the same INC_KEY identifiers. The PUF ICD Diagnosis data set included 3 variables: ICD CM diagnosis code, ICD CM diagnoses code Blank Inappropriate Values and ICD Clinical Modification version. This data set was used to distinguish TBI cases from cases related to other traumas such as pneumothorax, liver laceration or femur fractures. The way this was done was first the researcher identified TBI related ICD 10 CM diagnosis codes by visiting the ICD 10 Data website at www.icd10data.com and searching for all head injury related codes. Additionally,hydroponic drain table the selection of TBI related ICD 10 codes was corroborated by examining a list of codes found in existing studies on TBI which validated the inclusion of the specifically identified TBI codes in this study. Though these other studies included ICD 10 Diagnosis codes related to concussion injuries , these codes were excluded from this study as the researcher was only interested in identifying cases with either a moderate or severe TBI and concussions are designated as mild TBI. The following codes were ultimately selected: S02.0xx ; S02.1 ; S06.1 ; S02.19XD ;S06.2 ; S06.30 ; S06.31 ; S06.32 ; S06.33 ; S09.X . Next, PUF ICD 10 Diagnosis codes were regrouped into the following categories via numerical representation. ICD 10 Diagnosis code S02.0xx was grouped into group 3683-3687; S02.1 into group 3688; S02.19XD into group 3738; S06.1 into group 4008-4025; S06.2 into groups 4026-4045; S06.3, S06.31, S06.32, and S06.33 into groups 4046-4095; S09.X into groups 4310-4311. When missing values in each variable account for less than 5%, those values can be missing at random and list wise deletion can be performed relatively safely is appropriate to do. This holds true for all the variables except for THC Combo, positive for drugs, alcohol screen result, age in years, ethnicity and total GCS. These variables, three quantitative and three categorical, were found to have greater than 5% missing values. On observation of the missing value analysis, it was observed that most cases had these two variables as missing, perhaps suggesting a relationship, or an effect. Furthermore, the Little’s MCAR test revealed that missing data may not be missing completely at random. Deleting cases with missing values can reduce the statistical power of the analysis and result in biased outcomes and estimates.
Therefore, the use of multiple imputation is appropriate for this dataset and this study. Another method in SPSS that can be utilized is the Replacing Missing Values method. The Linear Interpolation method will be utilized. The Linear Interpolation method is a simple statistical method used by SPSS which estimates the value of one variable from the value of another and using regression methods to find the line of best fit. Using the Replacing Missing Values method in this study will help solve the problem of bias and ensure that power is not decreased because a large majority of the sample size will be preserved. It is important to consider the implications associated with imputing or replacing missing data. Multiple imputation or missing value replacement analyses will avoid bias only if enough variables predictive of missing values are included in the replacement method. If variables that may be predictive of the estimates are not included in the model, for example the effect of age on alcohol result, replacement computation will underestimate these associations and bias the final analysis. Therefore, it is preferrable to include as many predictive variables as possible in the model when either imputation or replacing missing value methods are utilized.Replacing missing values was utilized to minimize the many problems associated with missing data. The absence of data reduces statistical power and can also lead to bias in the estimation of parameters and analyses. Finally, missing data can diminish the representatives of the sample size and cases . It is important to consider that though replacing or imputing data is a common approach to the problem of missing data, it still does not allow analyses of actual data that is provided by actual participants, or in this case, data entered by abstractors and hospital registry systems. In gaining a larger sample size, and perhaps a more representative sample, confidence is lost that actual responses provided are those analyzed. It is important to note that methods used to account for missing data only provide researchers with the best estimated guess of what actual data may have been had it been documented in the first place. It is this ideology that influenced the decision to include some of the variables with missing data to be multiply imputed. Replacing missing values is another form of multiple imputation that was selected for this study. Though multiple imputation process was utilized, it presented a complication in terms of the number of iterations and the subsequent analysis.