Tag: statistics

PLoS One. 2023 Mar 13;18(3):e0282826. doi: 10.1371/journal.pone.0282826. eCollection 2023.

ABSTRACT

BACKGROUND: Nigeria has been consistently targeted in sub-Saharan Africa as an HIV-priority country. Its main mode of transmission is heterosexual, and consequently, a key population of interest is female sex workers (FSWs). While HIV prevention services are increasingly implemented by community-based organizations (CBOs) in Nigeria, there is a paucity of evidence on the implementation costs of these organizations. This study seeks to fill this gap by providing new evidence about service delivery unit cost for HIV education (HIVE), HIV counseling and testing (HCT), and sexually transmitted infection (STI) referral services.

METHODS: In a sample of 31 CBOs across Nigeria, we calculated the costs of HIV prevention services for FSWs taking a provider-based perspective. We collected 2016 fiscal year data on tablet computers during a central data training in Abuja, Nigeria, in August 2017. Data collection was part of a cluster-randomized trial examining the effects of management practices in CBOs on HIV prevention service delivery. Staff costs, recurrent inputs, utilities, and training costs were aggregated and allocated to each intervention to produce total cost calculations, and then divided by the number of FSWs served to produce unit costs. Where costs were shared across interventions, a weight proportional to intervention outputs was applied. All cost data were converted to US dollars using the mid-year 2016 exchange rate. We also explored the cost variation across the CBOs, particularly the roles of service scale, geographic location, and time.

RESULTS: The average annual number of services provided per CBO was 11,294 for HIVE, 3,326 for HCT, and 473 for STI referrals. The unit cost per FSW tested for HIV was 22 USD, the unit cost per FSW reached with HIV education services was 19 USD, and the unit cost per FSW reached by STI referrals was 3 USD. We found heterogeneity in total and unit costs across CBOs and geographic location. Results from the regression models show that total cost and service scale were positively correlated, while unit costs and scale were consistently negatively correlated; this indicates the presence of economies of scale. By increasing the annual number of services by 100 percent, the unit cost decreases by 50 percent for HIVE, 40 percent for HCT, and 10 percent for STI. There was also evidence that indicates that the level of service provision was not constant over time across the fiscal year. We also found unit costs and management to be negatively correlated, though results were not statistically significant.

CONCLUSIONS: Estimates for HCT services are relatively similar to previous studies. There is substantial variation in unit costs across facilities, and evidence of a negative relationship between unit costs and scale for all services. This is one of the few studies to measure HIV prevention service delivery costs to female sex workers through CBOs. Furthermore, this study also looked at the relationship between costs and management practices-the first of its kind to do so in Nigeria. Results can be leveraged to strategically plan for future service delivery across similar settings.

PMID:36913371 | DOI:10.1371/journal.pone.0282826

PLoS One. 2023 Mar 13;18(3):e0283045. doi: 10.1371/journal.pone.0283045. eCollection 2023.

ABSTRACT

BACKGROUND: Observational research on the advance care planning (ACP) process is limited by a lack of easily accessible ACP variables in many large datasets. The objective of this study was to determine whether International Classification of Disease (ICD) codes for do-not-resuscitate (DNR) orders are valid proxies for the presence of a DNR recorded in the electronic medical record (EMR).

METHODS: We studied 5,016 patients over the age of 65 who were admitted to a large, mid-Atlantic medical center with a primary diagnosis of heart failure. DNR orders were identified in billing records from ICD-9 and ICD-10 codes. DNR orders were also identified in the EMR by a manual search of physician notes. Sensitivity, specificity, positive predictive value and negative predictive value were calculated as well as measures of agreement and disagreement. In addition, estimates of associations with mortality and costs were calculated using the DNR documented in EMR and the DNR proxy identified in ICD codes.

RESULTS: Relative to the gold standard of the EMR, DNR orders identified in ICD codes had an estimated sensitivity of 84.6%, specificity of 96.6%, positive predictive value of 90.5%, and negative predictive value of 94.3%. The estimated kappa statistic was 0.83, although McNemar’s test suggested there was some systematic disagreement between the DNR from ICD codes and the EMR.

CONCLUSIONS: ICD codes appear to provide a reasonable proxy for DNR orders among hospitalized older adults with heart failure. Further research is necessary to determine if billing codes can identify DNR orders in other populations.

PMID:36913366 | DOI:10.1371/journal.pone.0283045

PLoS One. 2023 Mar 13;18(3):e0276419. doi: 10.1371/journal.pone.0276419. eCollection 2023.

ABSTRACT

Understanding the common topological characteristics of the human brain network across a population is central to understanding brain functions. The abstraction of human connectome as a graph has been pivotal in gaining insights on the topological properties of the brain network. The development of group-level statistical inference procedures in brain graphs while accounting for the heterogeneity and randomness still remains a difficult task. In this study, we develop a robust statistical framework based on persistent homology using the order statistics for analyzing brain networks. The use of order statistics greatly simplifies the computation of the persistent barcodes. We validate the proposed methods using comprehensive simulation studies and subsequently apply to the resting-state functional magnetic resonance images. We found a statistically significant topological difference between the male and female brain networks.

PMID:36913351 | DOI:10.1371/journal.pone.0276419

PLoS One. 2023 Mar 13;18(3):e0281981. doi: 10.1371/journal.pone.0281981. eCollection 2023.

ABSTRACT

The pandemic of COVID-19 is a severe threat to human life and the global economy. Despite the success of vaccination efforts in reducing the spread of the virus, the situation remains largely uncontrolled due to the random mutation in the RNA sequence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which demands different variants of effective drugs. Disease-causing gene-mediated proteins are usually used as receptors to explore effective drug molecules. In this study, we analyzed two different RNA-Seq and one microarray gene expression profile datasets by integrating EdgeR, LIMMA, weighted gene co-expression network and robust rank aggregation approaches, which revealed SARS-CoV-2 infection causing eight hub-genes (HubGs) including HubGs; REL, AURKA, AURKB, FBXL3, OAS1, STAT4, MMP2 and IL6 as the host genomic biomarkers. Gene Ontology and pathway enrichment analyses of HubGs significantly enriched some crucial biological processes, molecular functions, cellular components and signaling pathways that are associated with the mechanisms of SARS-CoV-2 infections. Regulatory network analysis identified top-ranked 5 TFs (SRF, PBX1, MEIS1, ESR1 and MYC) and 5 miRNAs (hsa-miR-106b-5p, hsa-miR-20b-5p, hsa-miR-93-5p, hsa-miR-106a-5p and hsa-miR-20a-5p) as the key transcriptional and post-transcriptional regulators of HubGs. Then, we conducted a molecular docking analysis to determine potential drug candidates that could interact with HubGs-mediated receptors. This analysis resulted in the identification of top-ranked ten drug agents, including Nilotinib, Tegobuvir, Digoxin, Proscillaridin, Olysio, Simeprevir, Hesperidin, Oleanolic Acid, Naltrindole and Danoprevir. Finally, we investigated the binding stability of the top-ranked three drug molecules Nilotinib, Tegobuvir and Proscillaridin with the three top-ranked proposed receptors (AURKA, AURKB, OAS1) by using 100 ns MD-based MM-PBSA simulations and observed their stable performance. Therefore, the findings of this study might be useful resources for diagnosis and therapies of SARS-CoV-2 infections.

PMID:36913345 | DOI:10.1371/journal.pone.0281981

Int J Surg. 2023 Feb 16. doi: 10.1097/JS9.0000000000000170. Online ahead of print.

ABSTRACT

OBJECTIVES: Post-prostatectomy urinary incontinence (PPUI) is a major complication that reduces the quality of life in patients undergoing prostatectomy for benign prostate hyperplasia (BPH) and prostate cancer (PCa). However, there are currently limited guidelines on which surgical techniques are preferred after conservative treatment for PPUI. In this study, a systematic review and network meta-analysis (NMA) that can help determine the priority for the selection of surgical methods were performed.

MATERIALS AND METHODS: We retrieved data from electronic literature searches of PubMed and the Cochrane Library through August 2021. We searched for randomized controlled trials (RCT) studies on the surgical treatment of PPUI after surgery for BPH or PCa and included the terms artificial urethral sphincter (AUS), adjustable sling, non-adjustable sling, and injection of the bulking agent.The NMA pooled the odds ratios (OR) and 95% credible intervals (CrI) using the number of patients achieving urinary continence, weight of pads used per day, number of pads used per day, and the International Consultation on Incontinence Questionnaire (ICIQ) score. The therapeutic effect of each intervention on PPUI was compared and ranked using the surface under the cumulative ranking curve (SUCRA).

RESULTS: A final 11 studies, including 1116 participants, were included in our NMA. The pooled overall ORs of patients achieving urinary continence compared with no treatment was 3.31 (95% CrI: 0.749, 15.710) in AUS, 2.97 (95% CrI: 0.412, 16.000) in adjustable sling, 2.33 (95% CrI: 0.559, 8.290) in non-adjustable sling, and 0.26 (95% CrI: 0.025, 2.500) in injection of bulking agent. In addition, this study shows the SUCRA values of ranking probabilities for each treatment performance, which indicated that AUS ranked first in terms of continence rate, ICIQ, pad weight, and pad use count.

CONCLUSION: The results of this study suggested that only AUS had a statistically significant effect compared to the non-treatment group and the highest PPUI treatment effect ranking among other surgical treatments.

PMID:36912884 | DOI:10.1097/JS9.0000000000000170

JMIR Form Res. 2023 Mar 13;7:e41408. doi: 10.2196/41408.

ABSTRACT

BACKGROUND: The use of information and communication technologies for health-eHealth-is described as having potential to improve the quality of health care service delivery. Consequently, there is an increased global trend toward adoption of eHealth interventions by health care systems worldwide. Despite the proliferation of eHealth solutions, many health care institutions especially in transitioning countries are struggling to attain effective data governance approaches. The Ministry of Health in Botswana is an exemplar institution continually seeking better approaches to strengthen health data governance (HDG) approaches following the adoption of eHealth solutions. Recognizing the need for a global HDG framework, the Transform Health coalition conceptualized HDG principles that are structured around 3 interconnected objectives: protecting people, promoting the value of health, and prioritizing equity.

OBJECTIVE: The aim of the study is to solicit and evaluate perceptions and attitudes of health sector workers in Botswana toward the HDG principles by Transform Health and derive any future guidance.

METHODS: Purposive sampling was used to select participants. A total of 23 participants from various health care organizations in Botswana completed a web-based survey and 10 participated in a follow-up remote round-table discussion. The aim of the round-table discussion was to gain further insight into participants’ responses from the web-based survey. Participants were from the following health care cadres: nurses, doctors, information technology professionals, and health informaticians. Both validity and reliability testing were performed for the survey tool before sharing it with study participants. An analysis of participants’ close-ended responses from the survey was performed using descriptive statistics. Thematic analysis of open-ended responses from the questionnaire and the round-table discussion was achieved using the Delve software and the widely accepted principles of thematic analysis.

RESULTS: Although some participants highlighted having measures in place similar to the HDG principles, there were some who either did not know or disagreed that their organizations already had in place mechanisms similar to the proposed HDG principles. Participants further expressed relevance and importance of the HDG principles in the context of Botswana. However, some modifications to the principles were also suggested.

CONCLUSIONS: This study highlights the necessity of data governance in health care particularly toward meeting the requirements for Universal Health Coverage. The existence of other health data governance frameworks calls for a critical analysis to assess the most appropriate and applicable framework in the context of Botswana and similar transitioning countries. An organization-centered approach may be most appropriate, as well as strengthening of existing organizations’ HDG practices with the Transform Health principles.

PMID:36912870 | DOI:10.2196/41408

Clin Orthop Relat Res. 2023 Mar 10. doi: 10.1097/CORR.0000000000002611. Online ahead of print.

ABSTRACT

BACKGROUND: THA for high-riding developmental dysplasia of the hip (DDH) is challenging in terms of length equalization. Although previous studies suggested preoperative templating on AP pelvic radiographs is insufficient in patients with unilateral high-riding DDH because of hypoplasia of the hemipelvis on the affected side and unequal femoral and tibial length on scanograms, the results were controversial. The EOS™ (EOS™ Imaging) is a biplane X-ray imaging system using slot-scanning technology. Length and alignment measurements have been shown to be accurate. We used the EOS to compare the lower limb length and alignment in patients with unilateral high-riding DDH.

QUESTIONS/PURPOSES: (1) Is there an overall leg length difference in patients with unilateral Crowe Type IV hip dysplasia? (2) In patients with unilateral Crowe Type IV hip dysplasia with an overall leg length difference, is there a consistent pattern of abnormalities in the femur or tibia that account for observed differences? (3) What is the impact of unilateral high-riding Crowe Type IV dysplasia on femoral neck offset and knee coronal alignment?

METHODS: Between March 2018 and April 2021, we treated 61 patients with THA for Crowe Type IV DDH (high-riding dislocation). EOS imaging was performed preoperatively in all patients. Eighteen percent (11 of 61) of the patients were excluded because of involvement of the opposite hip, 3% (two of 61) were excluded for neuromuscular involvement, and 13% (eight of 61) had previous surgery or fracture, leaving 40 patients for analysis in this prospective, cross-sectional study. Each patient’s demographic, clinical, and radiographic information was collected with a checklist using charts, Picture Archiving and Communication System, and an EOS database. EOS-related measurements that were related to the proximal femur, limb length, and knee-related angles were recorded for both sides by two examiners. The findings of the two sides were statistically compared.

RESULTS: The overall limb length was not different between the dislocated and nondislocated sides (mean 725 ± 40 mm versus 722 ± 45 mm, mean difference 3 mm [95% CI -3 to 9 mm); p = 0.08). Apparent leg length was shorter on the dislocated side (mean 742 ± 44 mm versus 767 ± 52 mm, mean difference -25 mm [95% CI -32 to 3 mm]; p < 0.001). We observed that a longer tibia on the dislocated side was the only consistent pattern (mean 338 ± 19 mm versus 335 ± 20 mm, mean difference 4 [95% CI 2 to 6 mm]; p = 0.002), but there was no difference between the femur length (mean 346 ± 21 mm versus 343 ± 19 mm, mean difference 3 mm [95% CI -1 to 7]; p = 0.10). The femur of the dislocated side was longer by greater than 5 mm in 40% (16 of 40) of patients and shorter in 20% (eight of 40). The mean femoral neck offset of the involved side was shorter than that of the normal side (mean 28 ± 8 mm versus 39 ± 8 mm, mean difference -11 mm [95% CI -14 to -8 mm]; p < 0.001). There was a higher valgus alignment of the knee on the dislocated side with a decreased lateral distal femoral angle (mean 84° ± 3° versus 89° ± 3°, mean difference – 5° [95% CI -6° to -4°]; p < 0.001) and increased medial proximal tibia angle (mean 89° ± 3° versus 87° ± 3°, mean difference 1° [95% CI 0° to 2°]; p = 0.04).

CONCLUSION: A consistent pattern of anatomic alteration on the contralateral side does not exist in Crowe Type IV hips except for the length of the tibia. All parameters of the limb length could be shorter, equal to, or longer on the dislocated side. Given this unpredictability, AP pelvis radiographs are not sufficient for preoperative planning, and individualized preoperative planning using full-length images of the lower limbs should be performed before arthroplasty in Crowe Type IV hips.

LEVEL OF EVIDENCE: Level I, prognostic study.

PMID:36912864 | DOI:10.1097/CORR.0000000000002611

JAMA Neurol. 2023 Mar 13. doi: 10.1001/jamaneurol.2023.0148. Online ahead of print.

ABSTRACT

IMPORTANCE: Pregnant women who have epilepsy need adequate engagement, information, and pregnancy planning and management to improve pregnancy outcomes.

OBJECTIVE: To investigate perinatal outcomes in women with epilepsy compared with women without epilepsy.

DATA SOURCES: Ovid MEDLINE, Embase, CINAHL, and PsycINFO were searched with no language or date restrictions (database inception through December 6, 2022). Searches also included OpenGrey and Google Scholar and manual searching in journals and reference lists of included studies.

STUDY SELECTION: All observational studies comparing women with and without epilepsy were included.

DATA EXTRACTION AND SYNTHESIS: The PRISMA checklist was used for abstracting data and the Newcastle-Ottawa Scale for risk-of-bias assessment. Data extraction and risk-of-bias assessment were done independently by 2 authors with mediation conducted independently by a third author. Pooled unadjusted odds ratios (OR) or mean differences were reported with 95% CI from random-effects (I2 heterogeneity statistic >50%) or fixed-effects (I2 < 50%) meta-analyses.

MAIN OUTCOMES AND MEASURES: Maternal, fetal, and neonatal complications.

RESULTS: Of 8313 articles identified, 76 were included in the meta-analyses. Women with epilepsy had increased odds of miscarriage (12 articles, 25 478 pregnancies; OR, 1.62; 95% CI, 1.15-2.29), stillbirth (20 articles, 28 134 229 pregnancies; OR, 1.37; 95% CI, 1.29-1.47), preterm birth (37 articles, 29 268 866 pregnancies; OR, 1.41; 95% CI, 1.32-1.51) and maternal death (4 articles, 23 288 083 pregnancies; OR, 5.00; 95% CI, 1.38-18.04). Neonates born to women with epilepsy had increased odds of congenital conditions (29 articles, 24 238 334 pregnancies; OR, 1.88; 95% CI, 1.66-2.12), neonatal intensive care unit admission (8 articles, 1 204 428 pregnancies; OR, 1.99; 95% CI, 1.58-2.51), and neonatal or infant death (13 articles, 1 426 692 pregnancies; OR, 1.87; 95% CI, 1.56-2.24). The increased odds of poor outcomes was increased with greater use of antiseizure medication.

CONCLUSIONS AND RELEVANCE: This systematic review and meta-analysis found that women with epilepsy have worse perinatal outcomes compared with women without epilepsy. Women with epilepsy should receive pregnancy counseling from an epilepsy specialist who can also optimize their antiseizure medication regimen before and during pregnancy.

PMID:36912826 | DOI:10.1001/jamaneurol.2023.0148

Liver Int. 2023 Mar 13. doi: 10.1111/liv.15561. Online ahead of print.

ABSTRACT

BACKGROUND & AIMS: Spleen stiffness measurement (SSM) by transient elastography (VCTE) has been tested in a limited number of studies versus hepatic venous pressure gradient (HVPG), especially with the 100 Hz spleen specific module. The current study aims to evaluate the diagnostic performance of this novel module for detecting clinically significant portal hypertension (CSPH) in a cohort of compensated patients with metabolic-associated fatty liver disease (MAFLD) as the main etiology and to improve the performance of the Baveno VII criteria for CSPH diagnosis by including SSM.

METHODS: This is a retrospective single center study including patients with available measurements of HVPG, liver stiffness (LSM) and SSM by VCTE with the 100 Hz module. AUROC analysis was conducted to identify dual cut-offs (rule-out and in) associated with the absence/presence of CSPH. The diagnostic algorithms were adequate if negative (NPV) and positive predictive values (PPV) were >90%.

RESULTS: A total of 85 patients were included, 60 MAFLD and 25 non-MAFLD. SSM showed a good correlation with HVPG (MAFLD: r=0.74; p<0.0001; non-MAFLD: r=0.62; p< 0.0011). In MAFLD patients, SSM had a high accuracy in discarding/diagnosing CSPH (cutoffs values of <40.9 and >49.9 kPa, AUC 0.95). The addition of these cutoffs in a sequential or combined approach to the Baveno VII criteria significantly reduced the grey zone (60% vs 15-20%), while maintaining adequate NPV and PPV.

CONCLUSIONS: our findings support the utility of SSM for diagnosing CSPH in MAFLD patients and demonstrates that the addition of SSM to the Baveno VII criteria increases accuracy.

PMID:36912787 | DOI:10.1111/liv.15561

Elife. 2023 Mar 13;12:e85082. doi: 10.7554/eLife.85082. Online ahead of print.

ABSTRACT

In this work, we expand the normative model repository introduced in (Rutherford, Fraza, et al., 2022) to include normative models charting lifespan trajectories of structural surface area and brain functional connectivity, measured using two unique resting-state network atlases (Yeo-17 and Smith-10), and an updated online platform for transferring these models to new data sources. We showcase the value of these models with a head-to-head comparison between the features output by normative modeling and raw data features in several benchmarking tasks: mass univariate group difference testing (schizophrenia versus control), classification (schizophrenia versus control), and regression (predicting general cognitive ability). Across all benchmarks, we show the advantage of using normative modeling features, with the strongest statistically significant results demonstrated in the group difference testing and classification tasks. We intend for these accessible resources to facilitate wider adoption of normative modeling across the neuroimaging community.

PMID:36912775 | DOI:10.7554/eLife.85082