Abstract
Background and Purpose Inspiratory muscle strength training (IMST) has been shown to improve maximal pressures and facilitate ventilator weaning in adults with prolonged mechanical ventilation (MV). The purposes of this case report are: (1) to describe the rationale for IMST in infants with MV dependence and (2) to summarize the device modifications used to administer training.
Case Description Two infants with congenital heart disease underwent corrective surgery and were referred for inspiratory muscle strength evaluation after repeated weaning failures. It was determined that IMST was indicated due to inspiratory muscle weakness and a rapid, shallow breathing pattern. In order to accommodate small tidal volumes of infants, 2 alternative training modes were devised. For infant 1, IMST consisted of 15-second inspiratory occlusions. Infant 2 received 10-breath sets of IMST through a modified positive end-expiratory pressure valve. Four daily IMST sets separated by 3 to 5 minutes of rest were administered 5 to 6 days per week. The infants' IMST tolerance was evaluated by vital signs and daily clinical reviews.
Outcomes Maximal inspiratory pressure (MIP) and rate of pressure development (dP/dt) were the primary outcome measures. Secondary outcome measures included the resting breathing pattern and MV weaning. There were no adverse events associated with IMST. Infants generated training pressures through the adapted devices, with improved MIP, dP/dt, and breathing pattern. Both infants weaned from MV to a high-flow nasal cannula, and neither required subsequent reintubation during their hospitalization.
Discussion This case report describes pediatric adaptations of an IMST technique used to improve muscle performance and facilitate weaning in adults. Training was well tolerated in 2 infants with postoperative weaning difficulty and inspiratory muscle dysfunction. Further systematic examination will be needed to determine whether IMST provides a significant performance or weaning benefit.
Mechanical ventilation (MV) is a signature therapy of modern intensive care and has saved countless lives. However, recent evidence has shown that the use of MV can rapidly lead to diaphragm muscle atrophy and weakness,1,2 termed “ventilator-induced diaphragm dysfunction” (VIDD). Because some modes of MV can dramatically reduce or stop diaphragm contractile activity, it is not surprising to observe VIDD following the use of MV. The effects of MV on diaphragm morphology and function include rapid onset of muscle fiber atrophy, reduced specific muscle tension, increased proteolysis, and decreased anabolic activity (for a review, see Powers et al3). It has been observed in animal models that the diaphragm atrophies approximately 8 times more rapidly than limb muscles.4 The animal research also has shown that even modest increases in diaphragm activity can attenuate VIDD compared with continuous MV support.5,6
Recent research on VIDD in humans indicates the extent of diaphragm atrophy and contractile dysfunction is similar in humans and smaller mammals.1,3,7 In adults who sustained critical traumatic injuries, diaphragm muscle fiber dimensions have been reported to be approximately 50% smaller within the first 68 hours of controlled MV.2 Diaphragmatic pressure decrements occur more slowly, but may diminish by one third within 6 days of MV introduction.8,9 Many improved critical care practices (sedation schedules, tight glycemic control, and optimized ventilator settings) do not directly address the coincident problem of VIDD and, therefore, may fail to result in extubation.10 In contrast to adults, VIDD is less well understood in children. However, specific diaphragm atrophy has been observed in mechanically ventilated infants and is thought to contribute to prolonged MV in children.11 Importantly, patients with VIDD and prolonged MV have a reduced ability to generate inspiratory pressure.3,8 The clinical implications of VIDD are profound because prolonged MV accompanies a high rate of serious comorbidities and a significantly higher risk of mortality.12
Encouragingly, ventilator weaning is significantly more successful for patients who improve their inspiratory pressure-generating capacity.13 Effective therapies for difficult ventilator weaning utilize the neuromuscular plasticity of the ventilatory muscles to prevent or reverse VIDD. Specific inspiratory muscle strength training (IMST) is a method that has been shown to enhance respiratory drive and activate ventilatory muscles in adults with respiratory insufficiency.14,15 Over the years, a number of small, uncontrolled studies demonstrated that IMST exercises increase maximal inspiratory pressure (MIP) and may facilitate ventilator weaning.16,17 Furthermore, a recent randomized trial showed a significant improvement in weaning outcome in patients who received IMST compared with sham training.18 Inspiratory muscle strength training is a promising therapeutic exercise for adults with weakness and failure to wean,15 but the effects of IMST in infants with VIDD are less well understood. Therefore, the purpose of this case report is to describe 2 infants who received IMST in conjunction with conventional strategies to facilitate postoperative weaning. We identify a potential use for IMST, detail the methods used to evaluate inspiratory muscle performance in infants, describe the equipment adaptations used to implement IMST in infants, and discuss the implications for further research.
Patient History and Systems Review: Case 1
Infant 1 was a female born at 37 weeks of gestation (birth weight=2.96 kg) and diagnosed with DiGeorge syndrome, a disorder characterized by gene deletion at the long arm of chromosome 22 (del22q11.2) and associated with palate and pharyngeal defects, hypoparathyroidism, thymus insufficiency, and learning disabilities.19 Although DiGeorge syndrome does not alter primary lung physiology, respiratory complications are common, due to impaired immune function and bronchopharyngeal structural deficits.20 In addition, approximately 40% of children with DiGeorge syndrome have a concurrent congenital heart defect.21
Infant 1 was diagnosed with truncus arteriosus (TA), a disorder that occurs when the pulmonary artery and aorta do not develop from a single, immature arterial trunk. This single vascular outlet receives blood from the right and left ventricles and delivers blood to the systemic, coronary, and pulmonary vascular systems. In addition to TA, the infant was found to have a patent ductus arteriosus, ventricular septal defect (VSD), and interrupted aortic arch. At 11 days of age, the infant underwent repair of the VSD, construction of a valved conduit between the pulmonary arteries and right ventricle, and anastomosis of the aorta. Her postoperative course was complicated by feeding intolerance, requiring gastric tube placement and Nissen fundoplication, as well as respiratory insufficiency and delayed extubation following each of the surgical interventions. She was originally discharged to home at 81 days of age.
Three days after her discharge to home, she was readmitted with increased gastric tube discharge and respiratory distress, and she was intubated upon arrival to the hospital (hospital day [HD] 1) (eTab. 1). During her hospitalization, the infant received treatment for pneumonia and management of the gastric wound. However, ventilator weaning was limited by pulmonary hypertension and left ventricular dysfunction due to significant pulmonary artery (PA) stenosis and a residual VSD. On HD 39, she underwent pulmonary valve replacement with a 9-mm allograft, VSD reclosure with a core matrix patch, right PA patch, right ventricular tract augmentation, and atrial catheter placement. The sternum was closed on HD 44. Despite optimization of her postoperative cardiac status, she failed 3 extubation trials over 10 days due to increased work of breathing and hypercapnic, hypoxemic respiratory failure. She required continuous MV support throughout the hospitalization. On HD 53 (patient age=147 days), a physical therapist was consulted to evaluate whether IMST was indicated for this patient. Her medical team felt that other pharmacological, surgical, and clinical therapies had been optimized, and parental consent was obtained for assessment and treatment. The primary goal of the surgeon and family was to facilitate ventilator weaning.
Patient History and Systems Review: Case 2
An infant boy born at 36 weeks (birth weight=2.22 kg) was diagnosed at birth with tetralogy of Fallot, double-outlet right ventricle (RV), pulmonary atresia, VSD, and aortopulmonary collaterals. He presented at age 49 days with poor feeding and worsening cyanosis (eTab. 2). He was intubated on HD 1 and underwent tetralogy of Fallot repair, RV-PA conduit, and VSD closure on HD 18. Postoperative ventilator weaning was impeded by left hemidiaphragm paralysis, and he underwent plication on HD 38. After plication, he developed acute cardiorespiratory failure requiring temporary extra-corporeal membranous oxygenation (ECMO) treatment from HD 48 to HD 55 and PA stent placement on HD 53. After ECMO discontinuation, he experienced repeated hypercapnia and hypoxemia during subsequent weaning attempts, and the physical therapist was consulted for an IMST evaluation on HD 63 (patient age=112 days). Parental consent was obtained, with a goal to optimize respiratory muscle function.
Initial Clinical Impression
The differential diagnosis included inspiratory muscle dysfunction and a progression of pre-existing congenital heart disease. We judged that IMST may be indicated for each infant due to: (1) a sustained decline from baseline respiratory function; (2) repeated hypercapnic respiratory failure, suggesting ventilatory muscle dysfunction; and (3) medical and nutritional stabilization. We selected MIP as the primary test to estimate inspiratory muscle strength. Although ventilatory muscle assessment of infants has been discussed in the literature,22,23 we found no previous quantitative reports of IMST used in children younger than school age. In each case, a parent provided consent to evaluate and treat the infant. The University of Florida does not require institutional review board approval for clinical case reports of 3 or fewer patients, but the infants' parents gave their written consent for a case report.
Examination
Inspiratory Muscle Performance
Reversible ventilatory muscle weakness is the primary indication for IMST exercises; therefore, our primary clinical evaluation was for strength. Inspiratory muscle strength was estimated noninvasively using MIP. The patients were tested in a reverse Trendelenburg position with approximately 30 degrees of head and trunk elevation. The patient was briefly disconnected from the ventilator, and then occlusion was provided by way of a unidirectional valve attached to the endotracheal tube, permitting exhalation (Fig. 1A). With each subsequent exhalation, the patient exhaled toward residual volume. The most negative pressure achieved in 15 seconds was recorded as the MIP (Fig. 2), and the best of 4 individual trials was recorded. The inspiratory occlusion maneuver has been determined to be a valid measure of inspiratory muscle strength in ventilated patients24 and found to yield maximal efforts in infants and young children.22,23 In addition, we calculated the rate of inspiratory pressure development (dP/dt) from the inspiratory attempt that generated the most negative pressure using the MIP divided by the time required to generate the pressure as an estimate of available neural drive.
Devices used to evaluate and train inspiratory muscle strength: (A) A unidirectional valve and respiratory monitor adapter was attached directly to the end of the endotracheal tube to measure maximal inspiratory pressure. This device also was used to provide inspiratory muscle strength training to infant 1. (B) An inverted positive end-expiratory pressure valve was used to provide a fixed pressure-threshold inspiratory load for infant 2's inspiratory muscle strength training.
Example of the pressure-time waveform generated during a 15-second inspiratory occlusion maneuver to evaluate maximal inspiratory pressure (MIP) in infant 2. The most negative pressure was the MIP, and rate of inspiratory pressure development was calculated from the time to generate MIP (red arrows).
Breathing Pattern
Breathing pattern was determined using a neonatal respiratory monitor (CO2SMO Plus with Capnostat neonatal adaptor, Philips-Respironics, Murrysville, Pennsylvania) connected to a laptop computer. We evaluated breathing pattern (inspiratory and expiratory times, inspired and expired volumes, peak inspiratory and expiratory flows, and negative inspiratory pressure) and end-tidal carbon dioxide levels while the infant was awake and calm, between 7:00 and 9:00 am. Dynamic compliance and resistance were calculated by the CO2SMO Plus using least-squares estimation. An intrinsic pneumotachograph and pressure transducer captured airflow and pressure at a rate of 100 Hz, and air flow was integrated to obtain volume. Data were processed with AnalysisPlus software (Philips-Respironics, Murrysville, Pennsylvania).
Clinical Impressions
In each case, echocardiograms indicated that cardiac function was stable at baseline. Additionally, complete blood count, basic metabolic profile, urine output, and chest radiographs did not reveal the presence of organ dysfunction or infection that would prohibit IMST. For both infants, our prediction of ventilatory muscle weakness was confirmed. Infant 1's initial MIP was −55.4 cm H2O, indicating a decrease of approximately 30% from age-referenced values, and the MIP of infant 2 was −31.7 cm H2O, a 60% reduction from age-referenced values.22
The inability to communicate directly with the infants was a precaution that required continuous monitoring to ensure the safety and tolerance of IMST. Moreover, the infants' inability to follow commands required a training mode that was not heavily influenced by motivation. However, the IMST device options were limited because commercially available devices have 30 to 40 mL of dead space and are intended for use in adults. In contrast, the infants' spontaneous tidal volumes were only approximately 20 mL.
Intervention
For infant 1, the MIP occlusion maneuvers were used for IMST. She was able to achieve 8 to 10 strong inspiratory efforts during MIP tests with stable oxygen saturation and heart rate. The combined dead space of our monitoring sensor and unidirectional valve was measured by volume displacement and was found to be approximately 5 mL. American Thoracic Society guidelines advise that the dead space of respiratory testing equipment optimally remain at ≤1.5 mL/kg of body mass.22 Because infant 1's endotracheal tube was uncuffed, small quantities of air leaked around the tube during occlusions, preventing the inspiratory muscle contractions from being truly “isometric.”
For infant 2, we used an inverted positive end-expiratory pressure (PEEP) valve and multiadaptor (15–22 mm) at the end of the endotracheal tube to provide pressure-threshold IMST (Accu-PEEP, Vital Signs Inc, Totowa, New Jersey). The combined dead space of the modified threshold device was approximately 15 mL (Fig. 1B). Pressure threshold devices consist of a one-way poppet valve closed by the tension of a spring. The valve requires patients to generate a minimum “threshold” pressure to overcome the spring tension, opening the valve and permitting air flow. Thus, threshold IMST involved an initial submaximal static occlusion, followed by dynamic respiratory muscle action during inhalation. We trained with the highest valve the infant could open completely and generate at least 50% of unloaded tidal volume.
For each case, IMST was conducted 5 to 6 mornings per week and consisted of four 15-second occlusions, each incorporating 8 to 12 inspiratory attempts, with at least 3 minutes of rest between occlusions. Respiratory parameters and vital signs were monitored continuously. The infants did not experience desaturation or increased end-tidal CO2 during or after training sessions. Inspiratory occlusions can elicit a pressor effect; therefore, particular attention was paid to blood pressure. Vital signs returned to baseline within approximately 3 to 5 minutes (eTabs. 3 and 4). We monitored the patients for at least 15 minutes after conclusion of IMST sessions.
In addition to IMST sessions, the patients continued the daily trials of reduced ventilator support that were initiated postsurgically. The duration and intensity of each sprint were determined by the medical and surgical physician team and typically involved sessions that were 1 to 2 hours long, with IMV or pressure support reduced 20% to 30% from baseline levels. Baseline MV pressure support and intermittent mandatory ventilation (IMV) levels were reassessed daily by the intensive care unit team and progressively reduced as spontaneous respiratory rate and tidal volumes improved.
The infants also received ongoing physical therapy management for positioning and gross motor developmental activities, 2 days per week. During the IMST interventional period, developmental therapies were scheduled in the later afternoon, after the children rested from IMST and trials of reduced ventilator support.
Outcome
The infants' tolerance to IMST was evaluated acutely by vital signs and adverse events. Medical stability was followed daily by review of radiology and laboratory reports, and over the course of treatment by echocardiogram. Clinical progress was tracked by MIP, dP/dt, breathing pattern, and ventilator weaning. Importantly, vital signs remained stable during and after IMST sessions. No adverse events, including acute infection, respiratory event (pneumonia, pleural effusion, pulmonary edema), or decline in organ function, occurred during IMST training days. Serial echocardiograms revealed existing valvular regurgitation, right ventricular hypertension, and pulmonary stenosis. Gas exchanging capacity (partial pressure of oxygen, arterial [PaO2]/fraction of inspired oxygen) was severely impaired in infant 2, but remained stable at the end of IMST (Table).
Resting Breathing Pattern and Respiratory Mechanicsa
Infant 1 participated in 13 IMST sessions over 15 days (HDs 54–68, patient age=148 days). Maximal inspiratory pressure increased 14% (from −55.4 cm H2O to −63.3 cm H2O, or from 47% to 53% of predicted values) (Fig. 3). The time to reach MIP decreased 27%. Thus, the infant was able to generate negative pressure more rapidly. Inspiratory dP/dt increased 43% (from 92 cm H2O/s to 132 cm H2O/s). In addition to strength gains, breathing pattern was improved from HD 53 to HD 68. Her baseline respiratory rate decreased 34%, and resting spontaneous tidal volumes increased nearly 60% (Table). After the 13th IMST session, the patient was extubated to a high-flow nasal cannula on HD 68. She was discharged to home using supplemental oxygen with a nasal cannula 2 weeks after extubation.
Inspiratory muscle performance of the infants: (A) maximal inspiratory pressure and (B) rate of inspiratory pressure development increased in both infants at the time of extubation. The “Pre” measurements reflect the infants' status at rest, prior to the initial evaluation. The “Post” measurements were taken at rest, prior to the last session of inspiratory muscle strength training and extubation.
Infant 2 participated in 5 IMST sessions over 7 days (HDs 64–70, patient age=112 days). Inspiratory muscle performance increased rapidly, and the training load increased from 7.5 cm H2O on the first training day to a final level of 10 cm H2O. The MIP on HD 63 was −31.7 cm H2O (27% of predicted values) and increased 76% by HD 70 to −56.0 cm H2O (47% of predicted values). With a 7% decrease in time to reach MIP, dP/dt increased 75%, from 72 H2O/s to 126 cm H2O/s (Fig. 3). Infant 2 experienced a 25% decrease in respiratory rate, and a 28% increase in spontaneous tidal volume during quiet breathing (Table). On HD 70, he was extubated to a high-flow nasal cannula and did not require reintubation. He was discharged to home 4 weeks later.
Discussion
Pediatric ventilator weaning failure is multifactorial and not well understood, yet the stakes remain especially high for infants and children. The risk of mortality increases 5-fold in children who experience extubation failures.25 Younger children and those with complex corrective surgeries are especially susceptible to difficult ventilator weaning or diaphragm paresis.26,27 We report 2 cases of improved MIP and breathing pattern in infants who tolerated IMST, accompanied by successful extubation. A number of factors may affect the ability of infants to wean from MV, including fluid balance, medications, cardiopulmonary function, and inspiratory muscle strength.25 In these 2 cases, fluid balance was stable, and blood urea nitrogen and creatinine remained at age-expected values. Scheduled analgesics and inotropic medications were unaltered during the IMST training period, and no concurrent corticosteroids or neuromuscular blocking agents were administered. Although each infant's medical comorbidities resulted in a significant failure to thrive (weight <10th percentile for age), nutritional management had been previously established by staff dietitians and was not acutely modified during our intervention. The chest radiograph and laboratory values were unchanged over the course of rehabilitation.
Special attention was paid to the potential cardiovascular effects of IMST. Negative thoracic pressure can increase cardiac preload (right-sided venous return), as well as afterload (left ventricular systolic pressure and myocardial workload). When ventilation shifts from mechanically assisted to unassisted breathing, the decreased intrathoracic pressure facilitates venous return to the right ventricle. However, even during large negative pressure fluctuations, venous return is flow limited, and these flow limitations reduce the potential for right ventricular overload during inspiratory occlusions.28
Inspiratory occlusion also may affect pulmonary vascular function or left ventricular work. Extreme negative thoracic pressures can raise pulmonary blood flow and elevate pulmonary pressure. On the other hand, hypoxemia and lung inflation have been found to influence pulmonary vascular resistance to a greater degree than intrathoracic pressure swings.29 Neither acute hypoxemia nor overinflation occurred during IMST sessions. Infants did not experience acute desaturation during the short training sets, and our monitoring equipment showed that unimpeded exhaled volumes exceeded the resisted IMST inhalations.
Large fluctuations in negative pressure also can transiently increase LV afterload and amplify myocardial oxygen consumption. Weekly echocardiograms were taken during training, including on the day prior to initiating IMST, and within 24 hours of extubation (eTabs. 5 and 6). Despite pre-existing congenital defects and ventricular functional impairments, no lasting shifts in the cardiopulmonary function of the infants occurred during the IMST period of the hospitalization.
The infants experienced comorbidities known to impair respiratory function, including pneumonia, delayed sternal closure, pulmonary hypertension, and diaphragm paresis. Each of these factors can independently prolong MV among infants who undergo cardiothoracic surgery.26 We did not find large fluctuations in dynamic compliance, but dynamic resistance decreased in both infants after the IMST period (Table; eTabs. 7 and 8). Because chest radiography, echocardiogram, and laboratory tests did not indicate appreciable changes in cardiopulmonary status, the reduced resistances may have been related to a decreased respiratory rate. Alternatively, the large air leak present in infant 1 is associated with overestimations of both compliance and resistance.30
The infant respiratory muscle pump is capable of generating a relatively high MIP because of the mechanical advantage described by LaPlace's law. The narrow diameter of the infant rib cage translates small quantities of muscular tension into relatively large pressures. Although infants who are healthy are capable of generating large MIP values, this capacity is counteracted by high minute ventilation and oxygen consumption requirements. Furthermore, intrinsic inspiratory loads consume a high proportion of capacity in infants compared with adults.22 The costovertebral angles approximate 90 degrees in newborns and minimize the contributions of intercostal muscles to tidal volumes.31 The infants described in this case report demonstrated both a decreased MIP and enhanced ventilatory loads due to high minute ventilation and metabolic requirements, as well as pre-existing elevated myocardial work. Physical therapists can contribute their knowledge of muscle physiology and biomechanics to interdisciplinary efforts to evaluate and optimize inspiratory muscle performance.
Weaning of patients from MV can involve a reduction of the programmed IMV rate or gradual reduction of inspiratory pressure support. In addition to IMST, the infants underwent daily weaning trials of reduced ventilator settings that may have functioned as a mild endurance exercise. Moreover, the baseline ventilator IMV settings were reduced progressively as the patients' resting respiratory rate slowed and spontaneous tidal volume increased.
Reductions in ventilator settings can accompany extubation readiness tests on minimal ventilatory support. However, weaning readiness indexes such as the rapid, shallow breathing index (RSBI=f/Vt, where f=frequency, and Vt=tidal volume) and the Compliance, Resistance, Oxygenation, Pressure Index (CROP Index=Cdyn × MIP × PaO2/Pao2 × f, where Cdyn=dynamic compliance, Pao2=partial pressure of oxygen, alveolar) have showed poor predictive value when applied to pediatric or adult patients.25,32 Likewise, strength alone does not reliably predict weaning success in individuals on MV, regardless of age.22 On the other hand, tension-time index is a highly significant predictor of weaning, both in adults and in infants and children.32,33 Successful MV weaning likely requires improvements in both respiratory strength and ventilatory timing. The patients progressively improved their strength and ventilatory pattern while receiving IMST and daily endurance “sprints.” However, we cannot isolate the role of IMST from the weaning and supportive care provided by the medical team, nor can we speculate upon specific neuromuscular adaptations that may have occurred with training. We urge caution when attempting to interpret the possible causes of successful weaning in these medically complex infants.
Few other data are available regarding the role of respiratory training in infants. A single published study of neonates demonstrated that light, flow-resistive inspiratory exercises increased endurance and improved resting tidal volume.34 In that study, peak airway pressures did not differ after training, but airway occlusion tests lasted for only 3 breaths while the infants were asleep. It is unlikely the investigators obtained maximal pressures using these brief occlusion tests. Nevertheless, the results suggested that training may have improved breathing pattern and performance of the neonates.
In conclusion, IMST was well-tolerated by 2 infants with cyanotic congenital heart disease and prolonged, postsurgical MV as part of an interdisciplinary weaning regimen. Two different modified devices were utilized to provide IMST. The infants increased their MIP over the training period and tolerated the exercise without complication. These results provide initial evidence that IMST may be feasible in some medically stabilized infants, but definitive evidence of its effectiveness in larger patient groups and value in weaning will require a controlled clinical trial.
Footnotes
All authors provided concept/idea/project design. Dr Smith and Dr Martin provided writing and data collection and analysis. Dr Bleiweis and Ms Neel provided patients, institutional liaisons, and consultation (including review of the manuscript before submission). Dr Martin provided facilities/equipment.
Case 1 of this case report was presented as a poster at the 2011 American Thoracic Society meeting; May 16, 2011; Denver, Colorado.
This work was supported by a Foundation for Physical Therapy Promotion of Doctoral Studies (PODS) II scholarship and National Institutes of Health predoctoral training grant support (NIH T32 HD043730) to Dr Smith.
The University of Florida and Dr Martin have applied for a patent to modify mechanical ventilators to provide threshold inspiratory muscle training to patients receiving mechanical ventilation.
- Received October 16, 2011.
- Accepted March 21, 2012.
- © 2013 American Physical Therapy Association
References
Issue highlights
