Print version ISSN 0034-7094
Rev. Bras. Anestesiol. vol.55 no.3 Campinas May/June 2005
Validation of a scale for the assessment of paravertebral muscle contraction during lumbar puncture*
Validación de escala de contracción de la musculatura paravertebral al estímulo de la punción lumbar
Tiago Gayer de Alencar, M.D.I; Gustavo Paiva Almeida, M.D.II; Gustavo Luchi Boos, M.D.III; Elizabeth Milla Tambara, TSA, M.D.IV; Getúlio Rodrigues de Oliveira Filho, TSA, M.D.V
IME1 do CET/SBA do Hospital
de Clínicas da UFPR, Curitiba, PR
IIME2 do CET/SBA Integrado de Anestesiologia da SES SC, Florianópolis, SC
IIIME3 do CET/SBA Integrado de Anestesiologia da SES SC, Florianópolis, SC
IVResponsável do CET/SBA do Serviço de Anestesiologia do Hospital de Clínicas da UFPR, Curitiba, PR
VResponsável do CET/SBA Integrado de Anestesiologia da SES SC, Florianópolis, SC
BACKGROUND AND OBJECTIVES: This study
aimed at validating a scale to quantify the intensity of paravertebral muscle
contraction during lumbar punctures, to be used in further studies on the effects
of local anesthetic infiltration.
METHODS: This study involved 31 physicians, from two different Anesthesiology training centers located in two Brazilian cities, 12 of which were from institution 1 and 19 from institution 2. Eight participants were first-year residents, 6 were second-year residents and 17 were instructors. Participants were invited to watch 23 movies, repeated in 3 randomized electronically generated samples. Movies presented skin stimulation with 25, 27 and 29 G Quincke needles during lumbar puncture and puncture simulations (syringe tip touch). Patients authorized shooting and showing the movies. Physicians evaluated paravertebral muscles contraction according to the following scale: 0 - no sign of muscle contraction; 1 - mild muscular contraction without lumbar spine movement; 2 - muscles contraction with moderate lumbar spine movement; 3 - muscle contraction with movement precluding needle progression. Modes of the three samples were considered for each movie score calculation. Reliability analysis consisted of calculation of Cronbach's alpha, and intercoder reliability was evaluated by Krippendorff's alpha coefficient.
RESULTS: Cronbach's alpha was 0.98. Median (25th - 75th percentiles) of Krippendorff's alpha coefficients was 0.81 (0.78 - 0.84).
CONCLUSIONS: High internal and intercoder reliability found in this study suggests that the scale may be used in future studies to quantify paravertebral muscle contraction during lumbar puncture.
Key words: MEASUREMENT TECHNIQUES: scale assessment paravertebral muscle
JUSTIFICATIVA Y OBJETIVOS: El objetivo
es validar escala para la cuantificación de la intensidad de la contracción
muscular paravertebral durante punciones lumbares, que auxiliará en la
evaluación de los efectos de la infiltración con anestésico
local en futuros trabajos.
MÉTODO: Participaron del estudio 31 observadores, de 2 Centros de Enseñanza y Entrenamiento (CEE-SBA), 12 del CEE-SBA 1 y 19 del CEE-SBA 2. Ocho participantes eran Médicos en Especialización (ME) de primer año, 6 de segundo año y 17 eran Instructores. A los evaluadores fueron mostradas 23 películas, repetidas en 3 muestras, según la secuencia aleatoria generada electrónicamente. Las películas contenían estímulo de la piel con agujas de Quincke 25, 27 y 29G durante punciones lumbares y simulaciones de punciones (toque de puntas de jeringas). Los pacientes autorizaron la filmación y utilización. Los movimientos de la musculatura paravertebral fueron cuantificados por los evaluadores, en la siguiente escala: 0 - ausencia de contracción visible; 1 - contracción leve, sin desplazamiento evidente de la columna; 2 - contracción, con movimentación moderada de la columna vertebral; 3 - contracción de la musculatura con movimentación que impide la progresión de la aguja. Para cálculo de los contajes de cada película, se consideraron las modas de las tres muestras. Se testó la consistencia interna por el coeficiente alfa de Cronbach y la concordancia entre los múltiples observadores por el coeficiente alfa de Krippendorff.
RESULTADOS: El coeficiente alfa de Cronbach fue 0,98. La mediana (cuartil inferior; superior) de los coeficientes alfa de Krippendorff fue 0,81 (0,78; 0,84).
CONCLUSIONES: La escala presentó un buen desempeño en lo que se refiere a la concordancia entre múltiples observadores, pudiendo ser utilizada para cuantificar la contracción de la musculatura paravertebral durante las punciones lumbares.
Some patients present paravertebral muscle contractions at lumbar puncture stimulation. This contraction may be explained by skin-muscle reflexes and their modulations 1,2. Paravertebral muscle contraction triggered by noxious stimulation may change spinal position thus increasing chances of failed spinal punctures 3. Since no scale was found in the literature to quantify the intensity of such contractions, and these data are intended to be used in future studies, a scale was developed for such quantification. Scale evaluation may be added to other methods of evaluating pain at lumbar puncture, such as visual analog and verbal scales.
So, this study aimed at validating a scale to quantify the intensity of paravertebral muscle contraction during lumbar punctures, to be used in further clinical studies to evaluate somatosensory responses to lumbar puncture.
This study was approved by the Ethics Committee, Hospital Governador Celso Ramos, Florianópolis, where movies were shot. Patients have consented and authorized movies shooting and showing. Patients were placed in the lateral position with broad lumbar region exposure. Movies were shot in the operating center with digital cameras before and during spinal anesthesia without preanesthetic medication. Content involved: puncture for skin infiltration with local anesthetics using hypodermal needles (0.45 x 13 mm); lumbar skin puncture with 25, 27 and 29G Quincke needles; and skin puncture simulations with needleless syringes.
Movies were evaluated by observers. Sample consisted of 31 observers between Specialization Physicians and Instructors from two different Anesthesiology training centers located in two Brazilian capital cities, 12 of which were from institution 1 and 19 from institution 2. From these, 8 were first year residents, 6 were second year residents and 17 were instructors.
Observers were presented with 23 lumbar puncture digital movies displayed in computer screens and repeated three times in randomized electronically generated sequence. A scale was given to observers to quantify paravertebral muscle movements during puncture of every movie they watched, with no influence of investigators.
Scale was graduated as follows:
0 - No sign of contraction;
1 - Mild contraction without vertebral column;
2 - Contraction with moderate vertebral column movement;
3 - Contraction with movements precluding needle progression.
Modes of the 3 samples were considered to calculate each observer's scores. Cronbach's alpha coefficient was used for internal reliability analysis 4 and Krippendorff's alpha coefficient was used for intercoder reliability analysis 5.
Krippendorff's alpha coefficient was calculated using the following formula 6:
a = 1 - (Do/De)
Do = 1 - 1/n (S Occ), where Occ = intercoder reliability, in a coincidence matrix.
De = 1 - 1/(n(n-1)) (S nc(nc-1)), where n = total number of observations and nc = number of observations in each category.
Cronbach's alpha was 0.98 and mean correlation among scores was 0.77. Median (25th - 75th percentiles) of Krippendorff's alpha coefficients was 0.81 (0.78 - 0.84). Krippendorff's alpha coefficient distribution is shown in figure 1.
Nociception is body perception and response to painful stimulations caused by complex nervous structures interactions. Cutaneous and dorsal painful stimulations activate free terminations of the axon of a primary afferent neuron in order to reach spinal cord. This stimulation may be transmitted by Ad and C fibers. Stimulation reaches spinal cord dorsal laminae and, after connecting to interneurons, stimulates spinal cord ventral horn where alpha motoneuron cellular bodies are located. Alpha motoneurons generate a stimulation transmitted to muscle fibers and promoting muscle contraction 2. Afferent stimulation passing through lumbar and thoracic spinal cord establishes neuronal connections which ascend by spinal cord and may suffer thalamic, cortical and descending pathways modulation. When stimulation comes from C fibers, modulation may be connected to gamma motoneuron connections which may change motor response. Action of these motoneurons may amplify the reflex, originating what is called descending multimodal spinal cord reflex2. This complex relation, starting with skin stimulation and ending with muscle contraction, is called skin-muscle reflex. This reflex is conditioned to a noxious stimulation since no effect was noticed with painless stimulations 1,2.
Our study has validated a scale for clinical quantification of paravertebral motor response to lumbar puncture. This motor response is induced by skin-muscle reflex presented by many patients submitted to such punctures and may lead to changes in patients' position thus impairing puncture 3.
For better practical use of this scale, it is important to highlight that all evaluation tools should be easy to apply and should have sufficient internal and intercoder reliability7. Internal reliability analysis (Cronbach's alpha) and intercoder reliability coefficients were used to evaluate such parameters4,5. In early evaluation stage, each observer has to agree with his/hers own evaluation in different moments. For this reason, we have exposed observers 3 times to each movie and final score was the mode among them.
In addition, evaluations have to be consistent among observers, because this way the scale may be reliably applied by different observers in different moments. For this purpose, intercoder reliability has to be calculated and there are several statistical methods for such. They may compare observers in pairs or, in a more complex way, they may compare more than two observers (multiple observers). Methods like Holsti, Scott, Spearman, Pearson and Cohen, among others, are used to pair observers. They differ among themselves for being more or less rigid in evaluations and results 5. However, it is suggested that the best way to compare intercoder reliability is Krippendorff's alpha coefficient 5.
Our study has calculated Krippendorff's alpha coefficient as from the table with all observers' evaluation because multiple observers were considered. It was first calculated by comparing each observer to all others (Figure 1) and then Krippendorff's alpha median for all observers was determined. For studies requiring the agreement of multiple observers, it is suggested that Krippendorff's alpha should be above 0.8, being accepted values above 0.7 5.
So, our conclusion was that the scale was adequate in terms of internal and intercoder reliability analysis 5, being a useful tool for clinical quantification of paravertebral muscle contraction during lumbar puncture.
01. Theriault E, Diamond J - Nociceptive cutaneous stimuli evoke localized contractions in a skeletal muscle. J Neurophysiol, 1988;60:446-462. [ Links ]
02. Zedka M, Prochazka A, Knight B et al - Voluntary and reflex control of human back muscles during induced pain. J Physiol, 1999;520:591-604. [ Links ]
03. de Filho GR, Gomes HP, da Fonseca MH et al - Predictors of successful neuraxial block: a prospective study. Eur J Anaesthesiol, 2002;19:447-451. [ Links ]
04. Bland JM, Altman DG - Cronbach's alpha. BMJ, 1997;314:572. [ Links ]
05. Lombard M, Snyder-Duch J, Bracken CC - Content analysis in mass communication - assessment and reporting of intercoder reliability. Human Communication Research, 2002;28:587-604. [ Links ]
06. http://www.temple.edu/mmc/reliability/out_hand_all.htm, acessado em 15/01/2005.
07. Bland JM, Altman DG - Validating scales and indexes. BMJ, 2002;324:606-607. [ Links ]
Dr. Tiago Gayer de Alencar
Address: Av. Presidente Kennedy, 811/31
ZIP: 80220-201 City: Curitiba, Brazil
Submitted for publication August 17, 2004
Accepted for publication February 9, 2005
* Received from CET/SBA do Hospital de Clínicas da Universidade Federal do Paraná (UFPR), Curitiba, PR