Pinch strength represents a fundamental clinical indicator for the assessment of hand function, particularly in rehabilitation, post-operative settings and biomechanical research. Its measurement is commonly performed using pinchmeters, among which the B&L Engineering pinch gauge is considered the gold standard. In recent years, Biometrics Ltd. has introduced the P100 pinchmeter, an electronic device characterised by a different design. Although this device has already been employed in some studies, the available literature has mainly focused on the evaluation of its reliability, without specifically assessing its concurrent validity. For this reason, the present study aims to analyse the concurrent validity of the Biometrics P100 with respect to the B&L Engineering gold standard, through an experimental protocol conducted at the Department of Bioengineering of Imperial College London. The study was conducted on a sample of 57 healthy adults, with a mean age of 25.58 years, recruited after providing informed consent and excluding subjects with neurological or musculoskeletal disorders affecting the hand. Each participant performed three types of pinch: two-point pinch (2P), three-point pinch (3P) and lateral pinch (LP), with both devices and with both hands. For each configuration, three consecutive repetitions were performed, separated by standardised contraction and recovery times, in order to limit muscle fatigue. To reduce possible bias, the order of the trials was randomised by means of an algorithm developed in MATLAB, while the posture of the participants was standardised in accordance with the guidelines of the American Society of Hand Therapists. The statistical analysis was performed after a preliminary assessment of data normality. Depending on the observed distribution, Spearman’s or Pearson’s correlation coefficients were applied, together with the Intraclass Correlation Coefficient (ICC), Bland-Altman analysis and the Wilcoxon signed-rank test or the paired t-test. The analysis was conducted according to two different approaches: the first based on the distinction between dominant and non-dominant hand, considering the maximum value obtained in the three repetitions; the second, adopted to allow direct comparison with the existing literature, based on the distinction between right and left hand and using the mean value of the three measurements. The results showed strong to very strong correlations between the two devices for all pinch types and for both hands, with coefficients ranging from 0.810 to 0.954 (p < 0.001). The ICC values, ranging from 0.749 to 0.891, indicated a level of agreement from good to excellent, with the highest values observed for LP. The Bland-Altman analysis confirmed an overall acceptable agreement between the two instruments, highlighting a slight systematic bias for 2P and LP, whereas for 3P the bias was negligible. Consistently with these results, the Wilcoxon signed-rank test and the paired t-test revealed statistically significant differences for 2P and LP, but not for 3P. The mechanical calibration performed using an Instron testing machine showed that the B&L PG-30 presents a non-linear response and an operator-dependent error of approximately 0.5 kg. On the contrary, the Biometrics P100 showed a more stable and reproducible behaviour, although presenting a slight tendency to overestimate. In conclusion, the Biometrics P100 demonstrated a satisfactory level of concurrent validity with respect to the B&L PG-30, proving to be a valid instrument for the measurement of pinch strength. However, the presence of systematic biases, although limited, in the 2P and LP tests suggests caution in the direct comparison of the two devices.
La forza di pinza rappresenta un indicatore clinico fondamentale per la valutazione della funzionalità della mano, soprattutto in ambito riabilitativo, post-operatorio e nella ricerca biomeccanica. La sua misurazione viene comunemente effettuata mediante pinchmeter, tra i quali il dispositivo considerato gold standard è il B&L Engineering pinch gauge. Negli ultimi anni, Biometrics Ltd. ha introdotto il pinchmeter P100, elettronico e caratterizzato da un design differente. Sebbene questo dispositivo sia già stato impiegato in alcuni studi, la letteratura disponibile si è concentrata prevalentemente sulla valutazione della sua affidabilità, senza valutarne in modo specifico la validità concorrente. Per questo motivo, il presente studio si propone di analizzare la validità concorrente del Biometrics P100 rispetto al gold standard B&L Engineering, attraverso un protocollo sperimentale condotto presso il Dipartimento di Bioingegneria dell'Imperial College London. Lo studio è stato condotto su un campione di 57 adulti sani, con età media pari a 25,58 anni, selezionati previo consenso informato ed esclusione di soggetti con patologie neurologiche o muscoloscheletriche a carico della mano. Ciascun partecipante ha eseguito tre tipologie di pinza: two-point pinch (2P), three-point pinch (3P) e lateral pinch (LP), con entrambi i dispositivi e con entrambe le mani. Per ogni configurazione sono state effettuate tre ripetizioni consecutive, intervallate da tempi standardizzati di contrazione e recupero, al fine di limitare l'affaticamento muscolare. Per ridurre possibili bias, l’ordine delle prove è stato randomizzato mediante un algoritmo sviluppato in MATLAB, mentre la postura dei partecipanti è stata standardizzata in accordo con le linee guida dell’American Society of Hand Therapists. L'analisi statistica è stata eseguita dopo una preliminare valutazione della normalità dei dati. A seconda della distribuzione osservata, sono stati applicati il coefficiente di correlazione di Spearman o di Pearson, l'Intraclass Correlation Coefficient (ICC), l'analisi di Bland-Altman e il test di Wilcoxon per i ranghi con segno o il paired t-test. L'analisi è stata condotta secondo due approcci differenti: il primo basato sulla distinzione tra mano dominante e non dominante, considerando il valore massimo ottenuto nelle tre ripetizioni; il secondo, adottato per consentire un confronto diretto con la letteratura esistente, basato sulla distinzione tra mano destra e sinistra e utilizzando il valore medio delle tre misurazioni. I risultati hanno mostrato correlazioni da forti a molto forti tra i due dispositivi per tutte le tipologie di pinza e per entrambe le mani, con coefficienti compresi tra 0,810 e 0,954 (p < 0,001). I valori di ICC, compresi tra 0,749 e 0,891, hanno indicato un livello di concordanza da buono ad eccellente, con i valori più alti osservati per la LP. L'analisi di Bland-Altman ha confermato un accordo complessivamente accettabile tra i due strumenti, evidenziando un bias sistematico lieve per 2P e LP, mentre per la 3P il bias è risultato trascurabile. Coerentemente con questi risultati, il test di Wilcoxon e il paired t-test hanno evidenziato differenze statisticamente significative per 2P e LP, ma non per la 3P. La calibrazione meccanica eseguita mediante macchina Instron ha mostrato che il B&L PG-30 presenta una risposta non lineare ed un errore dipendente dall'operatore di circa 0,5 kg. Al contrario, il Biometrics P100 ha evidenziato un comportamento più stabile e riproducibile, pur mostrando una lieve tendenza alla sovrastima. In conclusione, il Biometrics P100 ha dimostrato un livello soddisfacente di validità concorrente rispetto al B&L PG-30, risultando uno strumento valido per la misurazione della forza di pinza. Tuttavia, la presenza di bias sistematici, seppur contenuti, nelle prove di 2P e LP suggerisce cautela nel confronto diretto dei due dispositivi.
Validation of a commercially available pinch gauge and calibration with an uniaxial testing machine
De Luca, Lucrezia
2025/2026
Abstract
Pinch strength represents a fundamental clinical indicator for the assessment of hand function, particularly in rehabilitation, post-operative settings and biomechanical research. Its measurement is commonly performed using pinchmeters, among which the B&L Engineering pinch gauge is considered the gold standard. In recent years, Biometrics Ltd. has introduced the P100 pinchmeter, an electronic device characterised by a different design. Although this device has already been employed in some studies, the available literature has mainly focused on the evaluation of its reliability, without specifically assessing its concurrent validity. For this reason, the present study aims to analyse the concurrent validity of the Biometrics P100 with respect to the B&L Engineering gold standard, through an experimental protocol conducted at the Department of Bioengineering of Imperial College London. The study was conducted on a sample of 57 healthy adults, with a mean age of 25.58 years, recruited after providing informed consent and excluding subjects with neurological or musculoskeletal disorders affecting the hand. Each participant performed three types of pinch: two-point pinch (2P), three-point pinch (3P) and lateral pinch (LP), with both devices and with both hands. For each configuration, three consecutive repetitions were performed, separated by standardised contraction and recovery times, in order to limit muscle fatigue. To reduce possible bias, the order of the trials was randomised by means of an algorithm developed in MATLAB, while the posture of the participants was standardised in accordance with the guidelines of the American Society of Hand Therapists. The statistical analysis was performed after a preliminary assessment of data normality. Depending on the observed distribution, Spearman’s or Pearson’s correlation coefficients were applied, together with the Intraclass Correlation Coefficient (ICC), Bland-Altman analysis and the Wilcoxon signed-rank test or the paired t-test. The analysis was conducted according to two different approaches: the first based on the distinction between dominant and non-dominant hand, considering the maximum value obtained in the three repetitions; the second, adopted to allow direct comparison with the existing literature, based on the distinction between right and left hand and using the mean value of the three measurements. The results showed strong to very strong correlations between the two devices for all pinch types and for both hands, with coefficients ranging from 0.810 to 0.954 (p < 0.001). The ICC values, ranging from 0.749 to 0.891, indicated a level of agreement from good to excellent, with the highest values observed for LP. The Bland-Altman analysis confirmed an overall acceptable agreement between the two instruments, highlighting a slight systematic bias for 2P and LP, whereas for 3P the bias was negligible. Consistently with these results, the Wilcoxon signed-rank test and the paired t-test revealed statistically significant differences for 2P and LP, but not for 3P. The mechanical calibration performed using an Instron testing machine showed that the B&L PG-30 presents a non-linear response and an operator-dependent error of approximately 0.5 kg. On the contrary, the Biometrics P100 showed a more stable and reproducible behaviour, although presenting a slight tendency to overestimate. In conclusion, the Biometrics P100 demonstrated a satisfactory level of concurrent validity with respect to the B&L PG-30, proving to be a valid instrument for the measurement of pinch strength. However, the presence of systematic biases, although limited, in the 2P and LP tests suggests caution in the direct comparison of the two devices.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_07_DeLuca_Master_Thesis.pdf
solo utenti autorizzati a partire dal 25/06/2027
Descrizione: Tesi magistrale - De Luca
Dimensione
21.67 MB
Formato
Adobe PDF
|
21.67 MB | Adobe PDF | Visualizza/Apri |
|
2026_07_DeLuca_ExecutiveSummary.pdf
solo utenti autorizzati a partire dal 25/06/2027
Descrizione: Executive summary - De Luca
Dimensione
769.24 kB
Formato
Adobe PDF
|
769.24 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/260018