Is ChatGPT a trusted source of information for total hip and knee arthroplasty patients?

Aims

While internet search engines have been the primary information source for patients' questions, artificial intelligence large language models like ChatGPT are trending towards becoming the new primary source. The purpose of this study was to determine if ChatGPT can answer patient questions about total hip (THA) and knee arthroplasty (TKA) with consistent accuracy, comprehensiveness, and easy readability.

Methods

We posed the 20 most Google-searched questions about THA and TKA, plus ten additional postoperative questions, to ChatGPT. Each question was asked twice to evaluate for consistency in quality. Following each response, we responded with, "Please explain so it is easier to understand," to evaluate ChatGPT's ability to reduce response reading grade level, measured as Flesch-Kincaid Grade Level (FKGL). Five resident physicians rated the 120 responses on 1 to 5 accuracy and comprehensiveness scales. Additionally, they answered a "yes" or "no" question regarding acceptability. Mean scores were calculated for each question, and responses were deemed acceptable if ≥ four raters answered "yes."

Results

The mean accuracy and comprehensiveness scores were 4.26 (95% confidence interval (CI) 4.19 to 4.33) and 3.79 (95% CI 3.69 to 3.89), respectively. Out of all the responses, 59.2% (71/120; 95% CI 50.0% to 67.7%) were acceptable. ChatGPT was consistent when asked the same question twice, giving no significant difference in accuracy (t = 0.821; p = 0.415), comprehensiveness (t = 1.387; p = 0.171), acceptability (χ2 = 1.832; p = 0.176), and FKGL (t = 0.264; p = 0.793). There was a significantly lower FKGL (t = 2.204; p = 0.029) for easier responses (11.14; 95% CI 10.57 to 11.71) than original responses (12.15; 95% CI 11.45 to 12.85).

Conclusion

ChatGPT answered THA and TKA patient questions with accuracy comparable to previous reports of websites, with adequate comprehensiveness, but with limited acceptability as the sole information source. ChatGPT has potential for answering patient questions about THA and TKA, but needs improvement.

Clinician perspectives on nonoperative management of hip fractures during COVID-19

Aims

To identify factors influencing clinicians' decisions to undertake a nonoperative hip fracture management approach among older people, and to determine whether there is global heterogeneity regarding these factors between clinicians from high-income countries (HIC) and low- and middle-income countries (LMIC).

Methods

A SurveyMonkey questionnaire was electronically distributed to clinicians around the world through the Fragility Fracture Network (FFN)'s Perioperative Special Interest Group and clinicians' personal networks between 24 May and 25 July 2021. Analyses were performed using Excel and STATA v16.0. Between-group differences were determined using independent-samples t-tests and chi-squared tests.

Results

A total of 406 respondents from 51 countries answered the questionnaire, of whom 225 came from HIC and 180 from LMIC. Clinicians from HIC reported a greater median and mean estimated proportion of admitted patients with hip fracture undergoing surgery (median 96% (interquartile range (IQR) 95% to 99%); mean 94% (SD 8%)) than those from LMIC (median 85% (IQR 75% to 95%); mean 81% (SD 16%); p < 0.001). Global heterogeneity seems to exist regarding factors such as anticipated life expectancy, insufficient resources, ability to pay, treatment costs, and perception of risk in hip fracture management decision-making.

Conclusion

This study represents the first international sampling of clinician perspectives regarding nonoperative hip fracture management. Several factors seemed to influence the clinician decision-making process. Further research is needed to inform the development of best practice guidelines to improve decision-making and the quality of hip fracture care among older people.

Risk factors for unplanned reoperation after corrective surgery for adult spinal deformity

Aims

To determine the major risk factors for unplanned reoperations (UROs) following corrective surgery for adult spinal deformity (ASD) and their interactions, using machine learning-based prediction algorithms and game theory.

Methods

Patients who underwent surgery for ASD, with a minimum of two-year follow-up, were retrospectively reviewed. In total, 210 patients were included and randomly allocated into training (70% of the sample size) and test (the remaining 30%) sets to develop the machine learning algorithm. Risk factors were included in the analysis, along with clinical characteristics and parameters acquired through diagnostic radiology.

Results

Overall, 152 patients without and 58 with a history of surgical revision following surgery for ASD were observed; the mean age was 68.9 years (SD 8.7) and 66.9 years (SD 6.6), respectively. On implementing a random forest model, the classification of URO events resulted in a balanced accuracy of 86.8%. Among machine learning-extracted risk factors, URO, proximal junction failure (PJF), and postoperative distance from the posterosuperior corner of C7 and the vertical axis from the centroid of C2 (SVA) were significant upon Kaplan-Meier survival analysis.

Conclusion

The major risk factors for URO following surgery for ASD, i.e. postoperative SVA and PJF, and their interactions were identified using a machine learning algorithm and game theory. Clinical benefits will depend on patient risk profiles.

Cite this article: Bone Joint Res 2023;12(4):245–255.

Clinical outcomes after surgery for cervical radiculopathy performed in public and private hospitals : a nationwide relative effectiveness study

AIMS

The number of patients undergoing surgery for degenerative cervical radiculopathy has increased. In many countries, public hospitals have limited capacity. This has resulted in long waiting times for elective treatment and a need for supplementary private healthcare. It is uncertain whether the management of patients and the outcome of treatment are equivalent in public and private hospitals. The aim of this study was to compare the management and patient-reported outcomes among patients who underwent surgery for degenerative cervical radiculopathy in public and private hospitals in Norway, and to assess whether the effectiveness of the treatment was equivalent.

METHODS

This was a comparative study using prospectively collected data from the Norwegian Registry for Spine Surgery. A total of 4,750 consecutive patients who underwent surgery for degenerative cervical radiculopathy and were followed for 12 months were included. Case-mix adjustment between those managed in public and private hospitals was performed using propensity score matching. The primary outcome measure was the change in the Neck Disability Index (NDI) between baseline and 12 months postoperatively. A mean difference in improvement of the NDI score between public and private hospitals of ≤ 15 points was considered equivalent. Secondary outcome measures were a numerical rating scale for neck and arm pain and the EuroQol five-dimension three-level health questionnaire. The duration of surgery, length of hospital stay, and complications were also recorded.

RESULTS

The mean improvement from baseline to 12 months postoperatively of patients who underwent surgery in public and private hospitals was equivalent, both in the unmatched cohort (mean NDI difference between groups 3.9 points (95% confidence interval (CI) 2.2 to 5.6); p < 0.001) and in the matched cohort (4.0 points (95% CI 2.3 to 5.7); p < 0.001). Secondary outcomes showed similar results. The duration of surgery and length of hospital stay were significantly longer in public hospitals. Those treated in private hospitals reported significantly fewer complications in the unmatched cohort, but not in the matched cohort.

CONCLUSION

The clinical effectiveness of surgery for degenerative cervical radiculopathy performed in public and private hospitals was equivalent 12 months after surgery.Cite this article: Bone Joint J 2023;105-B(1):64-71.

The incidence and risk factors for stress fracture following periacetabular osteotomy

AIMS

The aims of this study were to characterize the incidence and risk factors associated with stress fractures following periacetabular osteotomy, and to determine their effect on osteotomy union.

METHODS

We retrospectively reviewed all periacetabular osteotomies (PAOs) performed for developmental dysplasia of the hip (DDH) at one institution over a six-year period between 2012 and 2017. Perioperative factors were recorded, and included demographic and surgical data. Postoperatively, patients were followed for a minimum of one year with anteroposterior and false profile radiographs of the pelvis to monitor for evidence of stress fracture and union of osteotomies. We characterized the incidence and locations of stress fractures, and used univariate and multivariable analysis to identify factors predictive of stress fracture and the association of stress fracture on osteotomy union.

RESULTS

A total of 331 patients underwent PAO during the study period with 56 (15.4%) stress fractures: 46 fractures of the retroacetabular posterior column, five cases of ischiopubic stress fracture, and five cases of concurrent ischiopubic and retroacetabular stress fractures. Overall, 86% (48/56) healed without intervention. Univariate analysis revealed that stress fractures occurred more frequently in females (p = 0.040), older patients (mean age 27.6 years (SD 8.4) vs 23.8 (SD 9.0); p = 0.003), and most often with the use of the broad Mast chisel (28.5%; p < 0.001). Multivariable analysis revealed that increasing age (odds ratio (OR) 1.04; 95% CI 1.01 to 1.07; p = 0.028) and use of the broad Mast chisel (OR 5.1 (95% CI 1.3 to 19.0) compared to narrow Ganz chisel; p = 0.038) and surgeon (p = 0.043) were associated with increased risk of stress fracture. Patients with stress fractures were less likely to have healed osteotomies after one-year follow-up (76% vs 96%; p < 0.001).

CONCLUSION

Stress fracture of the posterior column may be an under-recognized complication following PAO, and the rate may be influenced by surgical technique. Consideration should be given to using a narrow chisel during the ischial cut to reduce the risk of stress propagation through the posterior column.Cite this article: Bone Joint J 2022;104-B(9):1017-1024.

Testing of an Ultrasound-Limited Imaging Protocol for Pavlik harness Supervision (TULIPPS) in developmental dysplasia of the hip: a randomized controlled trial

AIMS

There is no consensus regarding optimum timing and frequency of ultrasound (US) for monitoring response to Pavlik harness (PH) treatment in developmental dysplasia of the hip (DDH). The purpose of our study was to determine if a limited-frequency hip US assessment had an adverse effect on treatment outcomes compared to traditional comprehensive US monitoring.

METHODS

This study was a single-centre noninferiority randomized controlled trial. Infants aged under six months whose hips were reduced and centred in the harness at initiation of treatment (stable dysplastic or subluxable), or initially decentred (subluxated or dislocated) but reduced and centred within four weeks of PH treatment, were randomized to our current standard US monitoring protocol (every clinic visit) or to a limited-frequency US protocol (US only at end of treatment). Groups were compared based on α angle and femoral head coverage at the end of PH treatment, acetabular indices, and International Hip Dysplasia Institute (IHDI) grade on one-year follow-up radiographs.

RESULTS

Overall, 100 patients were included; 42 patients completed the standard protocol (SP) and 40 completed the limited protocol (LP). There was no significant difference in mean right α angle at the end of treatment (SP 70.0° (SD 3.2°) ; LP 68.7° (SD 2.9°); p = 0.033), nor on the left (SP 69.0° (SD 3.5°); LP 68.1° (SD 3.3°); p = 0.128). There was no significant difference in mean right acetabular index at follow-up (SP 23.1° (SD 4.3°); LP 22.0° (SD 4.1°); p = 0.129), nor on the left (SP 23.3° (SD 4.2°); LP 22.8° (SD 3.9°); p = 0.284). All hips had femoral head coverage of > 50% at end of treatment, and all were IHDI grade 1 at follow-up. In addition, the LP group underwent a 60% reduction in US use once stable.

CONCLUSION

Our study supports reducing the frequency of US assessment during PH treatment of DDH once a hip is reduced and centred.Cite this article: Bone Joint J 2022;104-B(9):1081-1088.

Intraoperative pressure sensors improve soft-tissue balance but not clinical outcomes in total knee arthroplasty: a multicentre randomized controlled trial

AIMS

Intraoperative pressure sensors allow surgeons to quantify soft-tissue balance during total knee arthroplasty (TKA). The aim of this study was to determine whether using sensors to achieve soft-tissue balance was more effective than manual balancing in improving outcomes in TKA.

METHODS

A multicentre randomized trial compared the outcomes of sensor balancing (SB) with manual balancing (MB) in 250 patients (285 TKAs). The primary outcome measure was the mean difference in the four Knee injury and Osteoarthritis Outcome Score subscales (ΔKOOS₄) in the two groups, comparing the preoperative and two-year scores. Secondary outcomes included intraoperative balance data, additional patient-reported outcome measures (PROMs), and functional measures.

RESULTS

There was no significant difference in ΔKOOS₄ between the two groups at two years (mean difference 0.4 points (95% confidence interval (CI) -4.6 to 5.4); p = 0.869), and multiple regression found that SB was not associated with a significant ΔKOOS₄ (0.2-point increase (95% CI -5.1 to 4.6); p = 0.924). There were no significant differences between groups in other PROMs. Six-minute walking distance was significantly increased in the SB group (mean difference 29 metres; p = 0.015). Four-times as many TKAs were unbalanced in the MB group (36.8% MB vs 9.4% SB; p < 0.001). Irrespective of group assignment, no differences were found in any PROM when increasing ICPD thresholds defined balance.

CONCLUSION

Despite improved quantitative soft-tissue balance, the use of sensors intraoperatively did not differentially improve the clinical or functional outcomes two years after TKA. These results question whether a more precisely balanced TKA that is guided by sensor data, and often achieved by more balancing interventions, will ultimately have a significant effect on clinical outcomes. Cite this article: Bone Joint J 2022;104-B(5):604-612.

Changes and thresholds in the Forgotten Joint Score after total hip arthroplasty : minimal clinically important difference, minimal important and detectable changes, and patient-acceptable symptom state

AIMS

The aim of this study was to identify the minimal clinically important difference (MCID), minimal important change (MIC), minimal detectable change (MDC), and patient-acceptable symptom state (PASS) in the Forgotten Joint Score (FJS) according to patient satisfaction six months following total hip arthroplasty (THA) in a UK population.

METHODS

During a one-year period, 461 patients underwent a primary THA and completed preoperative and six-month FJS, with a mean age of 67.2 years (22 to 93). At six months, patient satisfaction was recorded as very satisfied, satisfied, neutral, dissatisfied, or very dissatisfied. The difference between patients recording neutral (n = 31) and satisfied (n = 101) was used to define the MCID. MIC for a cohort was defined as the change in the FJS for those patients declaring their outcome as satisfied, whereas receiver operating characteristic curve analysis was used to determine the MIC for an individual and the PASS. Distribution-based methodology was used to calculate the MDC.

RESULTS

Using satisfaction as the anchor, the MCID for the FJS was 8.1 (95% confidence interval (CI) 3.7 to 15.9; p = 0.040), which was affirmed when adjusting for confounding. The MIC for the FJS for a cohort of patients was 17.7 (95% CI 13.7 to 21.7) and for an individual patient was 18. The MDC90 for the FJS was eight, meaning that 90% of patients scoring more than this will have experienced a real change that is beyond measurement error. The PASS threshold for the FJS was defined as 29.

CONCLUSION

The MCID and MIC can be used respectively to assess whether there is a clinical difference between two groups, or whether a cohort or patient has had a meaningful change in their FJS. Both values were greater than measurement error (MDC90), suggesting a real change. The PASS threshold for the postoperative FJS can be used as a marker of achieving patient satisfaction following THA. Cite this article: Bone Joint J 2021;103-B(12):1759-1765.

Adverse childhood experiences and mental health in adulthood: Evidence from North Carolina

BACKGROUND

About 1.4 million adults in North Carolina (NC) have mental illnesses. Many get no treatment because of mental health care provider shortages. Necessary prevention requires identification of covariates of mental illness.

STUDY OBJECTIVE

We tested the hypotheses: There is no difference in adulthood mental health of the ACE-affected and the unaffected; ACEs do not significantly predict adulthood mental health.

METHODS

ACEs studied were: living with a mentally ill person, an alcoholic, or drug abuser; witnessing violence; having divorced or unmarried parents; and being physically, mentally or sexually abused. We used data from 2012 and 2014 NC Behavioral Risk Factor Surveillance System. We had 19,187 observations with about 13,900 including ACE data. We conducted t- and χ2 tests of mental health differences between ACE-affected and unaffected and regression analysis to determine mental health predictors.

RESULTS

ACE distribution was: 26.2% had parents or adults swearing at- or insulting them; 25% had divorced or unmarried parents; 22.6% lived with an alcoholic; 15.7% witnessed their parents beat each other; 13.8% lived with a mentally ill person; and 13.6% were hit or beaten by an adult in the home. Respondents also reported being touched sexually (9.4%); living with an adult who abused drugs (7.9%); and 4.1% were forced to have sex. The ACE-affected experienced significantly (p < 0.001) more days of poor mental health and had a higher likelihood of depression diagnosis than the unaffected. ACEs are statistically significant predictors of poor mental health in adulthood.

Medical and social factors as predictors of outcome in infant tracheostomy

We examined the relative impact of infant tracheostomy in comparison to associated medical and social factors, on developmental outcome as part of a cross-sectional follow-up of 32 children. These children had no mental retardation, physical handicap, or severe neurological problems, but had a history of long-term tracheostomy in infancy, ranging from 3 to 146 months duration. Medical factors evaluated included prematurity, neurological status, severity of illness, and number of weeks hospitalized. Social factors included parental education and occupation. Outcome measures included IQ, language quotient, growth parameters, and behavioral competence. Correlation analyses, stepwise multiple regression analyses, and t-tests were used. Early medical illnesses were significant predictors of cognitive, language, and growth outcome. Severity of medical complications at birth and the presence of any neurological problem predicted 49% of the variance in IQ at follow-up. Social class was the only variable to predict behavioral outcome, accounting for 28% of the variance. For children without confounding medical conditions, tracheostomy had a negative impact on overall language and auditory comprehension. Once children with confounding medical risk factors were removed from the sample, children with history of infant tracheostomy exhibited significantly lower overall mean language scores (106 versus 120), and lower mean language comprehension scores (104 versus 119) than a matched comparison group.