Missing data treatments matter: an analysis of multiple imputation for anterior cervical discectomy and fusion procedures

Ondeck, Nathaniel T.; Fu, Michael C.; Skrip, Laura; McLynn, Ryan P.; Cui, Jonathan J.; Basques, Bryce A.; Albert, Todd J.; Grauer, Jonathan N.

doi:10.1016/j.spinee.2018.04.001

Cited by 15 publications

(6 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One simulation study showed that missing data in more than 30% to 50% of cases can introduce bias. 25 The proportion of missing data in our study is similar to what was seen in a study assessing missing data on anterior cervical disk and fusion complication rates. 25 In this study, the authors used multiple imputation to replace missing data values of preoperative albumin and hematocrit, which constituted 63.5% of their study population.…”

Section: Discussionsupporting

confidence: 85%

“…25 The proportion of missing data in our study is similar to what was seen in a study assessing missing data on anterior cervical disk and fusion complication rates. 25 In this study, the authors used multiple imputation to replace missing data values of preoperative albumin and hematocrit, which constituted 63.5% of their study population. They then used their entire study cohort to show predictors of morbidity.…”

Section: Discussionsupporting

confidence: 85%

“…With a larger sample, it is possible that this variable would still have been associated with major complications even after excluding all patients with missing data, as it is possible that renal insufficiency does increase risk of certain complications. This not only shows the potential advantages of larger samples available in database studies 17 but also highlights the analytic benefits of the previously mentioned statistical techniques, 25 such as multiple imputation, to replace missing variables and have more robust sample sizes.…”

Section: Discussionmentioning

confidence: 71%

See 2 more Smart Citations

Impact of Missing Data on Identifying Risk Factors for Postoperative Complications in Hand Surgery

Aziz

Nayar

LaPorte

et al. 2021

Hand (New York, N,Y.)

View full text Add to dashboard Cite

Background The mismanagement of missing data in large clinical databases may lead to inaccurate findings. The purpose of this study was to demonstrate the effects of missing data on hand surgery research findings using an analysis of postoperative morbidity in patients undergoing hospital-based hand surgery. Methods The National Surgical Quality Improvement Program database was queried for patients undergoing common hand and upper extremity surgery between 2011 and 2016. Major and minor postoperative complications were identified. Demographics, comorbidity, and preoperative laboratory values were identified, and the percentage missing of each was tabulated. To demonstrate how missing data can alter analysis results, these variables were evaluated for an association with major complications using multivariable regression on 3 separate cohorts: (1) all patients; (2) all patients after exclusion of any patient entry with >10% of missing data; and (3) after removal of any patient entry with any missing data. Results Groups 1, 2, and 3 had 48 370, 23 118, and 6280 patients, respectively. There were 14 variables associated with increased odds of major complications in group 1, yet only 10 and 9 variables for groups 2 and 3, respectively. Six variables were associated with increased major complications across all 3 groups, whereas only 1 was associated with decreased odds of major complications across all groups. Conclusions Filtering patient cohorts according to the amount of missing patient information affected analyses of predictors for major complications associated with hospital-based hand surgery. These findings highlight the importance of considering and addressing missing data in large database studies.

show abstract

Section: Discussionsupporting

confidence: 85%

Section: Discussionsupporting

confidence: 85%

Section: Discussionmentioning

confidence: 71%

See 1 more Smart Citation

Impact of Missing Data on Identifying Risk Factors for Postoperative Complications in Hand Surgery

Aziz

Nayar

LaPorte

et al. 2021

Hand (New York, N,Y.)

View full text Add to dashboard Cite

show abstract

“…Use of the imputation of variables method has been used previously to account for missing variables and has been reported to eliminate potential selection bias that can arise from using complete case analysis. [41][42][43]…”

Section: Limitations and Strengthsmentioning

confidence: 99%

Preoperative Hyponatremia is a Risk Factor for Adverse 30-Day Outcomes Following Total Hip Arthroplasty

et al. 2020

View full text Add to dashboard Cite

The relationship between preoperative hyponatremia and 30-day outcomes following total hip arthroplasty (THA) is currently unknown. The present study used prospectively collected data to quantify the association between preoperative hyponatremia and odds of major morbidity (MM), longer length of stay, readmission, and reoperation within 30 days following THA. Patients who underwent THA between 2012 and 2014 were identified in the National Surgical Quality Improvement Program database using validated Current Procedural Terminology codes. Patients were included if they were either normonatremic or hyponatremic preoperatively. The outcome measures in this study were 30-day MM, hospital length of stay, 30-day readmission, and 30-day reoperation. A unique multivariable logistic regression model was used for each outcome to identify statistically significant associations between hyponatremia and the outcome of interest after adjusting for covariates. From 2012 to 2014, 59,236 THA procedures were recorded in National Surgical Quality Improvement Program, of which 55,611 patients were normonatremic and 3,051 patients were hyponatremic. After adjusting for covariates, preoperative hyponatremia was significantly associated with increased odds of MM (odds ratio [OR] = 1.14; 99% confidence interval [CI]: 1.01–1.30), 30-day reoperation (OR = 1.18; 99% CI: 1.02–1.36), and longer hospital length of stay (OR = 1.20; 99% CI: 1.13–1.27). Hyponatremia was not significantly associated with greater odds of 30-day readmission (OR = 0.91; 99% CI: 0.82–1.01). Preoperative hyponatremia was significantly associated with adverse 30-day outcomes following THA. As the U.S. health care system continues to transition toward value-based reimbursement that emphasizes health care quality, the results of the present study can be used to improve patient selection and preoperative counseling.

show abstract

“…24 Other limitations such as missing data within the database, as well as how a study reconciles this, are less well appreciated, although they can have profound effects on the study results. [25][26][27] In this issue of Arthroscopy, Aziz, Best, Shi, and Srikumaran 28 Using what is often considered a very highequality database, the National Surgical Quality Improvement Program (NSQIP) database, that has been shown to be superior to many administrative and commercial claims databases, the authors found that well over 90% of patients (93.5%) who underwent shoulder surgery had missing data and over half of these had a significant amount of missing data. The authors then analyzed various ways of dealing with the missing patient data and how it affected associations and complication risks after shoulder surgery, including the inclusion of all patients with any missing data (no exclusion), the exclusion of all patients with missing data from analysis, and the imputation of missing data using standardized means, among others.…”

Section: See Related Article On Page 1233mentioning

confidence: 99%

Editorial Commentary: Sometimes You Don’t Know What You’ve Got Until It’s Gone—The Effect of Missing Data in “Big Data” Studies

Dhawan¹

2020

Arthroscopy: The Journal of Arthroscopic & Related Surgery

View full text Add to dashboard Cite

Big-data studies are powerful tools for comparative-effectiveness research, but because of the large number of included patients, they risk falsely identifying a difference when none exists because large sample sizes may result in statistically significant differences that have little clinical importance. Other limitations of big-data studies include lack of generalizability because of inclusion of only specific patient populations, lack of validated outcome measures, recording bias or clerical error, and vast troves of missing data. As such, the methods and results of big-data studies require careful scrutiny to ensure that the conclusions are correct.

show abstract

Missing data treatments matter: an analysis of multiple imputation for anterior cervical discectomy and fusion procedures

Cited by 15 publications

References 30 publications

Impact of Missing Data on Identifying Risk Factors for Postoperative Complications in Hand Surgery

Impact of Missing Data on Identifying Risk Factors for Postoperative Complications in Hand Surgery

Preoperative Hyponatremia is a Risk Factor for Adverse 30-Day Outcomes Following Total Hip Arthroplasty

Editorial Commentary: Sometimes You Don’t Know What You’ve Got Until It’s Gone—The Effect of Missing Data in “Big Data” Studies

Contact Info

Product

Resources

About