A multicenter study of interocular symmetry of corneal biometrics in Chinese myopic patients

AbstractIt is essential to know the normal range of the interocular symmetry of the cornea (ISC) for keratoconus diagnosis and corneal substitutes design. In the present study we investigated the interocular symmetry of corneal biometrics in 6,644 Chinese myopic patients from multiple ophthalmic centers. Corneal biometrics of both eyes were exported from the Pentacam instrument. Interocular symmetry of the corneal biometrics was analyzed by Spearman’s correlation test, intraclass correlation coefficient (ICC) analysis and Bland–Altman plot. Significantly strong interocular correlations were found in anterior and posterior corneal curvatures, corneal diameter, corneal thickness, corneal volume, corneal eccentricity, and corneal asphericity (r = 0.87–0.98, all P < 0.001). Moderate interocular correlations were observed in whole corneal astigmatism (r = 0.78) and posterior corneal astigmatism (r = 0.73). ICC between the right and left eyes was 0.94–0.98 for anterior and posterior corneal curvatures, corneal diameter, corneal thickness and corneal volume, 0.80–0.88 for corneal eccentricity and asphericity, and 0.73–0.79 for corneal astigmatism (all P < 0.001). Bland–Altman plot showed that the 95% limit of agreement between both eyes was narrow and symmetric in most of the corneal biometrics, suggesting strong interocular agreements in these corneal biometrics. In conclusion, significant interocular symmetry of corneal biometrics is observed in Chinese myopia patients. Extra attention should be paid to patients with interocular corneal asymmetry.

Download Full-text

Do the Readings of Digital and Aneroid Sphygmomanometer Concur? A Clinic-Based Study in an Urban Area of South Kolkata

JOURNAL FOR CLINICAL AND DIAGNOSTIC RESEARCH ◽

10.7860/jcdr/2021/48122.15178 ◽

2021 ◽

Author(s):

Aparajita Dasgupta ◽

Foulisa Pyrbot ◽

Bobby Paul ◽

Soumit Roy ◽

Pritam Ghosh ◽

...

Keyword(s):

Blood Pressure ◽

Blood Pressure Measurement ◽

Intraclass Correlation ◽

Research Work ◽

P Value ◽

Altman Plot ◽

Cohen’S Kappa ◽

Predictive Values ◽

Bland Altman Plot ◽

Cohen's Kappa

Introduction: Hypertension is a major risk factor for cardiovascular and cerebrovascular diseases. Thus, regular and accurate measurement of Blood Pressure (BP) is essential for its early diagnosis and follow-up. There is a surge in popularity of digital sphygmomanometer due to its convenience of use and functionality. In contrast, the traditional universally accepted sphygmomanometer is aneroid type, hence there arise a need for comparison of digital and universally accepted Aneroid Sphygmomanometer in terms of agreement and correlation. Aim: To evaluate the agreement and correlation between blood pressure measurement by digital and aneroid sphygmomanometer. Materials and Methods: The clinic based cross-sectional study was conducted in the Out Patient Department (OPD) of Urban Heath Centre, All India Institute of Hygiene and Public Health, Kolkata, West Bengal, India. Adults visiting the OPD on two chosen days of the week, between June 2019 to July 2019 were selected using systematic random sampling. A total of 400 participants were included. Agreement and correlation between BP measurements by digital and aneroid sphygmomanometer was analysed by Cohen’s Kappa, Bland Altman Plot along with sensitivity, specificity and predictive values using Microsoft Excel and Statistical Package for the Social Sciences (SPSS) version 16.0. the p-value <0.05 was considered significant for the statistical test in the analysis. Results: Cohen’s Kappa value (0.59) revealed these two tools had moderate agreement in diagnosing hypertension. Sensitivity and specificity of digital sphygmomanometer taking aneroid sphygmomanometer as gold standard is 86% and 83.1% respectively. The BP readings of these two-tools showed moderate correlation as Intraclass Correlation Coefficient (ICC) for Systolic BP (SBP) and Diastolic BP (DBP) were 0.804 and 0.624, respectively. Bland Altman plot showed gross disagreement of SBP findings and disagreement between DBP findings was also noted. Conclusion: Digital device was found to be less accurate in detecting hypertension. Therefore, more similar research work is solicited to verify the accuracy of the very easy to use, the Digital BP monitor.

Download Full-text

Comparisons of Five Equations Used for Estimating the Glomerular Filtrations Rate in Chinese Adults

10.21203/rs.3.rs-17896/v1 ◽

2020 ◽

Author(s):

shuai fu ◽

Min Yang ◽

Si Xu ◽

Sha Wu ◽

Xiao Xiao ◽

...

Keyword(s):

Intraclass Correlation ◽

Roc Curves ◽

Severely Injured ◽

Chinese Adults ◽

Altman Plot ◽

Bland Altman Plot ◽

Chinese Adult ◽

Blood Biochemical ◽

Hospitalised Patients ◽

Kappa Value

Abstract Background We aimed to assess the performance of revised MDRD, CKD-EPI, BIS, FAS and XiangYa equation in Chinese adults Methods We collected blood biochemical data of 623 chinese adult hospitalised patients within 48 hours before they underwent 99m Tc-DTPA GFR measurement. We computed the bias (mGFR-eGFR), the precision (IQR) ,the accuray (P30）and root mean square error (RMSE) relative to mGFR of each equation to evaluate performance. The ROC curves, Kappa value of McNemar test, Bland-Altman plot and the Intraclass correlation coefficient (ICC) were used to evaluate diagnostic accuracy and concordance. Results Totally, the FAS combined Scr and cysC equation performed supreme accuracy(P30=57.5%, RMSE=19.26), the cysC-based equation performed superior to Scr-based equation. Detailed P30 of the CKD-EPI cysC , FAS cysC , MDRD, CKD-EPI Scr-cysC , CKD-EPI Scr , FAS Scr , XiangYa was 56.7%, 56.0%, 53.5%, 52.2%, 48.8%, 51.4%, 43.0%. The CKD-EPI cysC equation showed the lowest bias and the highest accuracy(bias=-2.23, P30=57.4%) in GFR<60ml/min/1.73m 2 , followed by the FASscr-cysC equation(bias=-6.89, P30=55.4%). The XiangYa equation perfomed best in GFR≥60ml/min/1.73m 2 while worst in GFR<60 ml/min/1.73m 2 with bias(-5.79 vs -19.05), IQR(18.21 vs 10.85), P30(86.2% vs 21.1%), RMSE(16.68 vs 21.34). The CKD-EPI cysC equation had the lowest bias and the best accuracy(bias=-2.23, P30=59.4%) in age ≥70 years adults, followed by the FAS Scr-cysC equation equivalented to BIS-2 Scr-cysC equation(bias -5.33 vs -4.90, P30=57.3%), while the XiangYa equation performed worstly (bias=-20.39, P30=26.6%). Best ROC AUC was gaven by the FAS Scr-cysC equation(0.951),so was it had the highest Kappa value(0.364). The lowest Bias showed in Bland-Altman plot was the CKD-EPI cysC equation(bias=7.46). The highest ICC value was gaven by the FAS Scr-cysC equation(0.921). Secondly, it was the XiangYa equation with the ICC of 0.912. Conclusions The FAS Scr-cysC equation is verified most suitable and simpler applied to Chinese population. The CKD-EPI cysC equation is appropriate used in moderately and severely injured GFR(CKD3-5stage) and Seniors over 70 years old. The XiangYa equation performed perfectly in slightly injured GFR (CKD1-2stage), while further verification of XiangYa equation in multiple region need to carried out especially in moderately and severely injured GFR and older adults.

Download Full-text

In VivoCorneal Biomechanical Properties with Corneal Visualization Scheimpflug Technology in Chinese Population

BioMed Research International ◽

10.1155/2016/7840284 ◽

2016 ◽

Vol 2016 ◽

pp. 1-9 ◽

Cited By ~ 2

Author(s):

Ying Wu ◽

Lei Tian ◽

Yi-fei Huang

Keyword(s):

Linear Regression ◽

Healthy Subjects ◽

Chinese Population ◽

Regression Models ◽

Corneal Thickness ◽

Intraclass Correlation ◽

Biomechanical Properties ◽

Linear Regression Models ◽

Corneal Volume ◽

Biomechanical Parameters

Purpose. To determine the repeatability of recalculated corneal visualization Scheimpflug technology (CorVis ST) parameters and to study the variation of biomechanical properties and their association with demographic and ocular characteristics.Methods. A total of 783 healthy subjects were included in this study. Comprehensive ophthalmological examinations were conducted. The repeatability of the recalculated biomechanical parameters with 90 subjects was assessed by the coefficient of variation (CV) and intraclass correlation coefficient (ICC). Univariate and multivariate linear regression models were used to identify demographic and ocular factors.Results. The repeatability of the central corneal thickness (CCT), deformation amplitude (DA), and first/second applanation time (A1/A2-time) exhibited excellent repeatability (CV%≤3.312% and ICC≥0.929 for all measurements). The velocity in/out(Vin/out), highest concavity- (HC-) radius, peak distance (PD), and DA showed a normal distribution. Univariate linear regression showed a statistically significant correlation betweenVin,Vout, DA, PD, and HC-radius and IOP, CCT, and corneal volume, respectively. Multivariate analysis showed that IOP and CCT were negatively correlated withVin, DA, and PD, while there was a positive correlation betweenVoutand HC-radius.Conclusion. The ICCs of the recalculated parameters, CCT, DA, A1-time, and A2-time, exhibited excellent repeatability. IOP, CCT, and corneal volume significantly influenced the biomechanical properties of the eye.

Download Full-text

Correlation of Corneal Thickness Measurements Between Tono-pachymeter and Scheimpflug- Placido Topography

Medicine & Health ◽

10.17576/mh.2021.1602.12 ◽

2021 ◽

Vol 16 (2) ◽

pp. 161-167

Author(s):

Zeynep Eylül Ercan ◽

Keyword(s):

Corneal Thickness ◽

Age Groups ◽

Limits Of Agreement ◽

Altman Plot ◽

Bland Altman Plot ◽

Significant Relationships ◽

Plot Analysis ◽

Group 2 ◽

Thickness Measurements ◽

Group 1

Central corneal thickness (CCT) measurements are important for diagnosis, treatment, and surgery planning in ophthalmology. The purpose of this study was to see whether CCT measurements taken with Tono-pachymeter and Scheimpflug- Placido Topography had any significant differences. Tono-pachymeter and topography CCT measurements were taken (n=400). Inter-measurement agreement between them was determined using Bland-Altman Plot analysis. Age groups were also formed as group 1 (aged 18-50 years, 94 males, 106 females) and group 2 (age >51 years, 100 males, 100 females). Mean CCTs measured by Tonopachymeter and topography were 563.77 +±26.43 and 560.88 + 26.341 microns. Bland-Altman Plot analysis showed in total, 13 were above the upper limit and 5 were under the minimum limit of agreement with regression analysis showing no significant relationships (p=0.213). Group 1 had 7 above and 2 below from the limits of agreement. Group two had 9 above and 2 below from the limits of agreement. Both groups showed insignificant differences between devices (p=0.07 and p=0.86). Tono-pachymeter and Scheimpflug-Placido Topography give reliable CCT results within each other. However, since the limit of agreement ranges can still affect one-to-one patient evaluations, we recommend clinics that use these devices to not interchange measurements in practice.

Download Full-text

A Comparative Study of the Saltzman and the Modified Long-leg Radiograph for Assessing Hindfoot Alignment

Foot & Ankle Orthopaedics ◽

10.1177/2473011417s000364 ◽

2017 ◽

Vol 2 (3) ◽

pp. 2473011417S0003

Author(s):

Kalpesh Shah ◽

Kumar Kaushik Dash

Keyword(s):

Correlation Coefficient ◽

Intraclass Correlation Coefficient ◽

Knee Replacement ◽

Lower Limb ◽

Intraclass Correlation ◽

Limb Length Discrepancy ◽

Altman Plot ◽

Linear Distribution ◽

Hindfoot Alignment ◽

Bland Altman Plot

Category: Radiology Introduction/Purpose: There is growing evidence suggesting that the alignment of the leg (hip-knee-ankle or HKA axis) should be considered when planning an ankle replacement and alignment of the hindfoot (tibio-calcaneal or TC axis) should be considered when planning a knee replacement. The alignment of the HKA axis is assessed from standard long-leg radiographs, while the hindfoot alignment requires special views like Saltzman’s or Coetzee’s or similar. However we have discovered that by making minor adjustments to the way a standard long-leg radiograph is taken, it is possible to measure both the HKA as well as the TC axis from the same radiograph. The aim of our study was to compare the hindfoot alignment between a Saltzman’s view and the modified long-leg view. Methods: Approval was obtained for a prospective observational study on 65 consecutive patients referred to our hospital for a knee replacement. Patients with multiple joint arthritis, previous lower-limb surgery, or limb length discrepancy requiring shoe- raise were excluded. All patients had a preoperative modified long-leg view as part of our routine for radiographic assessment of a knee replacement; as well as a Saltzman’s view which served as a standard for hindfoot alignment. The long-leg alignment was measured using the standard HKA (hip-knee-ankle) axis, and the alignment of the hindfoot (tibio-calcaneal angle) was measured using the method described by Van Dijk et all. A difference of 3 degrees between the two observations was agreed to be considered as significant. Results were recorded by two Surgeons independently and analysed using the Bland-Altman Plot. The Intraclass Correlation Coefficient was used for inter-observer variation. Results: 48 patients were eligible for the study. The hindfoot angle measured in the modified long-leg view (mean 9.29 degrees, SD 4.26 degrees) was similar to that in the Saltzman’s view (mean 8.99 degrees, SD 4.16 degrees). The Scatter Plot showed a linear distribution, with Spearmann Correlation Coefficient of 0.892. Bland Altman Plot showed the differences in the measurements residing within the agreed difference of 3 degrees. Intraclass Correlation Coefficient was 0.94 for the modified long-leg view and 0.87 for the Saltzman’s view, suggesting excellent interobserver agreement. Conclusion: Our study shows that the position of the hindfoot can be reliably measured from the modified long-leg view, and this removes the need for additional radiographs for hindfoot alignment. As the entire femur, tibia and the hindfoot is accessible in the same long-leg view, this creates a platform for further research in to the mechanical axis of the entire lower limb as opposed to the HKA axis, and the effects of malalignment of the lower extremity in patients undergoing knee or ankle replacements.

Download Full-text

Comparison of central corneal thickness and intraocular pressure measured with two different tono/pachymeter devices in non-glaucomatouse children

European Journal of Ophthalmology ◽

10.1177/1120672121991492 ◽

2021 ◽

pp. 112067212199149

Author(s):

Kuddusi Teberik ◽

Mehmet Tahir Eski ◽

Handan Ankarali

Keyword(s):

Intraocular Pressure ◽

Central Corneal Thickness ◽

Corneal Thickness ◽

Intraclass Correlation ◽

Applanation Tonometry ◽

Healthy Children ◽

Clinical Trial Registration ◽

Registration Number ◽

The Mean ◽

The Right

Purpose: The goal of this research is to compare the intraocular pressure (IOP) and the mean central corneal thickness (CCT) values obtained from the measurements with Nidek NT-530P and Canon TX-20P devices with the values obtained by ultrasound pachymetry (UP) and Goldmann applanation tonometry (GAT) in children. Methods: This prospective study was conducted with 119 healthy children. The measurements were repeated three times for each eye. The intraclass correlation coefficient (ICC) was used to assess the correlation between the measurements obtained from different devices. The Bland–Altman plot was used to analyze the agreement between two different devices graphically. The measurements were taken in the same order in all subjects: TX-20P, NT-530P, UP, and GAT. Results: The mean age of the children was 10.1 ± 3.2 (6–17) years. The mean CCT values for the eyes were 568.90 and 569.68 (TX-20P), 571.44 and 566.37 (NT-530P), 564.77 and 564.67 μm UP, (right and left, respectively). The highest correspondence observed for CCT was between UP and NT-530P devices (ICC, 0.982, the left eye). The mean IOP measurements for Canon TX-20P, NT-530P, and GAT were 16.5 ± 3.2, 16.3 ± 3.2, and 16.8 ± 3.6 mmHg for the right eyes, respectively, while those for the left eyes were 16.6 ± 3.1, 16.2 ± 3.3, and 16.8 ± 3.7 mmHg. The highest correspondence was noted between measurements obtained using GAT and Nidek NT-530P devices (ICC, 0.945, in the left eye). Conclusion: Tono/pachymetry devices in the healthy children are provides significant agreements with the gold standard methods for IOP and CCT measurements. Clinical Trial Registration number: 2016-94

Download Full-text

07 Examination of Reliability and Validity of Index of Postural Stability for Patients with Neurological Diseases

Age and Ageing ◽

10.1093/ageing/afz164.07 ◽

2019 ◽

Vol 48 (Supplement_4) ◽

pp. iv1-iv2

Author(s):

Hisashi Mochizuki

Keyword(s):

Neurological Disease ◽

Postural Stability ◽

Neurological Diseases ◽

Intraclass Correlation ◽

Reliability And Validity ◽

Minimal Detectable Change ◽

Altman Plot ◽

Bland Altman Plot ◽

Functional Reach Test ◽

Balance Ability

Abstract Purpose The purpose of this study was to examine the reliability and validity of the Index of Postural Stability for patients with neurological disease. Methods The Index of postural stability (IPS) is a logarithmic value of the ratio between the stability limit and the area of centre of pressure (COP) in standing position, and is based on a probabilistic idea. Subjects were 75 patients with neurological disease (mean age; 63±12 years, 47 with Parkinson's disease, 9 with multiple sclerosis, 8 with Parkinsonism and others 11). Balance ability of subjects were assessed using IPS, Berg balance scale (BBS), Timed Up and Go Test(TUG), Functional Reach Test(FR), and gait ability of subjects were assessed using Functional Gait Category(FAC). IPS was assessed twice at intervals of about 30 minutes. Reliability of IPS was examined using Intraclass correlation coefficient (ICC), Bland-Altman plot and minimal detectable change (MDC). Validity of IPS was examined using correlation coefficients with BBS, TUG, FRT, and FAC. Results IPS showed significant and moderate correlation with BBS (r = 0.54), FR (r = 0.36), TUG (r =-0.43). IPS and FAC showed strong correlation (rs= 0.64). ICC of 2 separate measurements was 0.96. From the results of Bland-Altman plot, IPS had no additional error or proportional error (95%CI:-0.048~0.034, r=-0.0088, p=0.94). The MDC of IPS was 0.34. Discussion Since IPS has high reproducibility and does not have additional errors or proportional errors, and is also significantly related to other evaluation indexes of balance ability, it is considered that IPS may be a useful evaluation index of balance ability for patients with neurological diseases.

Download Full-text

Original article. Repeatability assessment of the New Numbers Contrast Sensitivity Chart

Asian Biomedicine ◽

10.5372/1905-7415.0803.300 ◽

2014 ◽

Vol 8 (3) ◽

pp. 361-369

Author(s):

Bharkbhum Khambhiphant ◽

Sunee Chansangpetch ◽

Wasee Tulvatana ◽

Mathu Busayarat

Keyword(s):

Contrast Sensitivity ◽

Narrow Range ◽

Primary Language ◽

Altman Plot ◽

Bland Altman Plot ◽

Plot Analysis ◽

Good Repeatability ◽

The Mean ◽

Mean Differences ◽

The Right

Abstract Background: The validity and agreement of the New Numbers Contrast Sensitivity Chart with the original Mars chart have been found to be good. The two charts can be used interchangeably; however, evidence of the repeatability of the new chart remains to be sought. This study was to assess the repeatability of the New Numbers Contrast Sensitivity Chart. Objectives: We assessed the repeatability of the Numbers Contrast Sensitivity Chart. Methods: Two hundred subjects from the ophthalmic clinic of the King Chulalongkorn Memorial Hospital, who were able to communicate and read Arabic numerals were recruited. The contrast sensitivity (CS) scores were collected by reading the same Numbers Contrast Sensitivity Chart in the same environment with each eye and both eyes ten minutes apart. The repeatability of the CS score was assessed by Bland-Altman plot analysis. Results: The visual acuity of subjects with variety in diagnoses ranged from 20/480 to 20/20. The mean differences were -0.006, -0.008, -0.002 log CS and the coefficients of repeatability were 0.155, 0.141, and 0.093 for the right eye, left eye, and both eyes, respectively. The plots showed a narrow range of 95% limit of agreement, which were (+0.146, -0.159) in the right eye, (+0.130, -0.147) in the left eye, and (+0.089, -0.093) in both eyes. Conclusions: The New Numbers Contrast Sensitivity Chart has good repeatability. With proven good validity and repeatability, this easy and convenient numbers chart is beneficial for practical use in a clinical setting where English is not used as the primary language.

Download Full-text

Comparing face-to-face and videoconference completion of the Montreal Cognitive Assessment (MoCA) in community-based survivors of stroke

Journal of Telemedicine and Telecare ◽

10.1177/1357633x19890788 ◽

2019 ◽

pp. 1357633X1989078 ◽

Cited By ~ 3

Author(s):

Jodie E Chapman ◽

Dominique A Cadilhac ◽

Betina Gardner ◽

Jennie Ponsford ◽

Ruchi Bhalla ◽

...

Keyword(s):

Repeated Measures ◽

Cognitive Assessment ◽

Intraclass Correlation ◽

Screening Tools ◽

Montreal Cognitive Assessment ◽

Cognitive Screening ◽

Community Based ◽

Altman Plot ◽

Bland Altman Plot ◽

Face To Face

Introduction Videoconferencing may help address barriers associated with poor access to post-stroke cognitive screening. However, the equivalence of videoconference and face-to-face administrations of appropriate cognitive screening tools needs to be established. We compared face-to-face and videoconference administrations of the Montreal Cognitive Assessment (MoCA) in community-based survivors of stroke. We also evaluated whether participant characteristics (e.g. age) influenced equivalence. Methods We used a randomised crossover design (two-week interval). Participants were recruited through community advertising and use of a stroke-specific database. Both sessions were conducted by the same researcher in the same location. Videoconference sessions were conducted using Zoom. A repeated-measures t-test, intraclass correlation coefficient (ICC), Bland–Altman plot and multivariate regression modelling were used to establish equivalence. Results Forty-eight participants (26 men, Mage = 64.6 years, standard deviation ( SD) = 10.1; Mtime since stroke = 5.2 years, SD = 4.0) completed the MoCA face-to-face and via videoconference on average 15.8 ( SD = 9.7) days apart. Participants did not perform systematically better in a particular condition, and no participant variable predicted difference in MoCA performance. However, the ICC was low (0.615), and the Bland–Altman plot indicated wide limits of agreement, indicating variability between sessions. Discussion Our findings provide preliminary evidence to support the use of videoconference to administer the MoCA following stroke. However, further research into the test–retest reliability of scores derived from the MoCA is needed in this population. Administering the MoCA via videoconference holds potential to ensure that all stroke survivors undergo cognitive screening, in line with recommended clinical practice.

Download Full-text

Level of agreement between sPRO and Kubios software in the analysis of R-R intervals obtained by a chest strap

Proceedings of the Institution of Mechanical Engineers Part P Journal of Sports Engineering and Technology ◽

10.1177/17543371211031145 ◽

2021 ◽

pp. 175433712110311

Author(s):

José Pino-Ortega ◽

Markel Rico-González ◽

Petrus Gantois ◽

Fabio Y Nakamura

Keyword(s):

Time Domain ◽

Time Course ◽

Intraclass Correlation ◽

Systematic Bias ◽

Altman Plot ◽

Bland Altman Plot ◽

Software Packages ◽

The Mean ◽

Improved Performance ◽

Level Of Agreement

Heart rate variability (HRV) is becoming one of the most commonly used tools for tracking the time course of training adaptation/maladaptation of athletes and setting optimal training loads, leading to improved performance. The aim of this study was to compare Realtrack Systems manufacturer software (sPRO) with the Kubios HRV Standard. R-R intervals were recorded from 48 athletes using a chest strap synchronized to WIMU PRO™ device for 10 min to analyze time-domain HRV indices, located between the scapulae. Time-domain HRV indices analyzed included the square root of the mean squared differences of successive R-R intervals (RMSSD), standard deviation of normal R-R intervals (SDNN), and percentage of adjacent R-R intervals that differ from each other by more than 50 ms (pNN50). The intraclass correlation test and Bland-Altman plot were used to verify the agreement between the sPRO and the Kubios HRV Standard software for examining time-domain HRV indices. The intraclass correlation coefficient showed high values of agreement of R-R intervals and time-domain HRV indices between the two software options (RMSSD, SDNN, and pNN50; ICC > 0.951). The Bland-Altman plot showed low bias for all the HRV indices analyzed (bias = −0.11 to −1.08). In addition, no systematic bias was found between the residual of the mean difference and the average values of the two software packages ( p > 0.05). Therefore, both software options can be used interchangeably for analyzing time-domain HRV records.

Download Full-text