Planning Military Drinking Water Needs: Development of a User-Friendly Smart Device Application

Transcription

1 MILITARY MEDICINE, 181, 9:1142, 2016 Planning Military Drinking Water Needs: Development of a User-Friendly Smart Device Application Nisha Charkoudian, PhD*; Robert W. Kenefick, PhD*; Anthony J. Lapadula, PhD ; Albert J. Swiston, PhD ; Tajesh Patel, BS ; Laurie A. Blanchard, BS ; Elizabeth M. Caruso, MS*; Anthony J. Luippold, BS*; Samuel N. Cheuvront, PhD* ABSTRACT Potable water is essential to maintain health and sustain military operations, but carrying and transporting water is a major logistical burden. Planning for group drinking water needs is complex, requiring understanding of sweat losses on the basis of intensity of activity, clothing biophysical parameters, and environmental conditions. Use of existing prediction equations is limited to tabled doctrine (e.g., Technical Bulletin, Medical 507) or to individuals with extensive expertise in thermal biophysics. In the present project, we translated the latest updated equations into a user-friendly Android application (Soldier Water Estimation Tool, SWET) that provides estimated drinking water required from 5 simple inputs based upon a detailed multiparametric sensitivity analysis. Users select from multiple choice inputs for activity level, clothing, and cloud cover, and manually enter exact values for temperature and relative humidity. Total drinking water needs for a unit are estimated in the Mission Planner tool on the basis of mission duration and number of personnel. In preliminary user acceptability testing, responses were overall positive in terms of ease of use and military relevance. Use of SWET for water planning will minimize excessive load (water) carriage in training and mission settings, and will reduce the potential for dehydration and/or hyponatremia to impair Warfighter health and performance. INTRODUCTION Drinking water is the largest single potable water planning factor for the U.S. Armed Forces, and represents 15% of the total per capita water planning. 1 Provision of adequate water is a challenge on the modern battlefield, especially in mountainous terrains and/or when roads are poorly maintained or nonexistent. In such cases, water is either carried on foot or air-dropped, which requires tremendous manpower, vehicle space, and fuel consumption, making water transport one of the largest military logistical supply burdens. From a medical and physiological perspective, lack of adequate water results in dehydration, which increases risk of serious heat illness and performance impairment, especially during prolonged exposure to wet bulb globe temperatures above 30 C. 2 5 Conversely, overhydration is the principal cause of a potentially fatal condition known as hyponatremia, or water intoxication. 6 8 Thus, when water is not available or is incorrectly managed, combat effectiveness quickly declines. The key to planning for appropriate water consumption for groups of soldiers is the ability to accurately predict *Thermal and Mountain Medicine Division, U.S. Army Research Institute of Environmental Medicine, General Greene Avenue, Natick, MA MIT - Lincoln Laboratory, 244 Wood Street, Lexington, MA Biophysics and Biomedical Modeling Division, U.S. Army Research Institute of Environmental Medicine, General Greene Avenue, Natick, MA The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Army or the Department of Defense. Citations of commercial organizations and trade names in this report do not constitute an official Department of the Army endorsement or approval of the products or services of these organizations. Approved for public release; distribution unlimited. doi: /MILMED-D water needs across a range of environmental conditions, with varying activity levels, loads, clothing ensembles, and personal protective equipment. Accurate knowledge of sweat losses is critical, particularly for active populations and when exposed to heat stress. Currently, military potable water planning relies on water tables generated from measurements made during a variety of laboratory and field experiments. 2,8,9 Estimates of water and electrolyte needs for more varied conditions and more prolonged environmental exposures 2 have relied upon a legacy sweat prediction equation 10 that overpredicts water needs in many circumstances. 11 To address the need for improved prediction accuracy, Gonzalez et al developed and validated an updated algorithm based on a piece-wise (PW) model that better predicts observed sweat losses (i.e., drinking water needs) in cool, temperate, warm-hot, high-altitude, and transient solar load environments This new algorithm is based on metabolic demands, clothing biophysical parameters, and environmental conditions like the legacy equation, 10 but expands the range of relevant environments, metabolic rates, and mission durations, and includes modern clothing ensembles (e.g., body armor) to provide estimates of water needs in both training and operational scenarios. However, use of this updated algorithm is limited to individuals with extensive training in thermal biophysics and ultimately requires numerous measured inputs. For this reason, tabled doctrine, such as that published in Technical Bulletin, Medical has been generated to provide training guidance. The military user community requested greater simplicity and flexibility for predicting water needs rapidly and in real time particularly during operational tempo training. The goal of the present project was to incorporate the updated water planning algorithm into a smartphone/smart 1142

2 device application (app) that would provide important, accurate mission planning information in a rapid, easy-to-access format. The resulting product is the Soldier Water Estimation Tool (SWET), an app for Android smartphones. The primary target user for the current version of the SWET app is military unit leaders responsible for planning group drinking water logistics for mission or training activities. Nett Warrior (PEO Soldier, Ft. Belvoir, Virginia) is an integrated dismounted leader situational awareness system, including a set of apps, for use during combat operations of the U.S. Army. In April 2015, SWET was fielded within the Nett Warrior system as one of its first external apps. The purpose of this manuscript is to describe the development and testing of SWET, leading up to its fielding by PM Nett Warrior earlier this year. METHODS Platform We translated the PW model from a Microsoft Excelbased (Microsoft Corporation, Redmond, Washington) format (provided by Dr. RR Gonzalez) into an Android smartphone app platform. Although several mobile platforms were considered, the Android platform was chosen as it has been mandated for use by the U.S. Army for mobile app development, and is consistent with requirements for future development and fielding (e.g., by the Nett Warrior program). Although the resulting app should execute well on any modern Android phone, the present testing was performed primarily on the Samsung Galaxy Note II and the Samsung Galaxy Nexus smartphones (Samsung Electronics America, Ridgefield Park, New Jersey). The current prototype is being used on Samsung Galaxy Note II devices running both stock Android operating systems as well as the Army s Nett Warrior platform. Simplification of Input Variables (Sensitivity Analysis) A further goal of this effort was to identify the variables from the original model that were the strongest contributors to sweat rate estimation to minimize the number of necessary inputs. The accuracy of the published algorithm is ±125 ml/ h. 13 This error, referred to as the zone of indifference, is based on the desire to predict individual sweat losses to within 1 L of actual water needs over an 8-hour workday. 11 Numerous variables (>35) are required inputs to the PW model. Several variables were identified, for which a single constant value could be substituted without exceeding the fluid predictions margin of error ( ±125 ml/h). This helped to reduce the number of potential user inputs to 11. We then conducted a sensitivity analysis via parameter sweeps of the 11 remaining parameters, which allowed further simplification to five input parameters as a result of natural crosscorrelations between interactive parameters. For example, body weight has a direct influence on metabolic rate, but it has no separate influence on sweating if metabolic rate is held constant. As a result, if metabolic rate is entered correctly, other, related variables such as body weight can be held constant. The 5 required inputs are discussed below. Metabolic Rate A parameter sweep was performed for activity level from 250 to 600 watts (W), which represents a range of military activity from weapon maintenance and drill to field assaults. The change in sweat range estimation output was well above the 125 ml/h minimum level, so metabolic rate was included in the final parameter set. However, since small changes in activity level do not have a substantial impact on sweating rate, the final input provides three categorical selections: easy (250 W), moderate (425 W), and heavy work (600 W). Each of these choices includes an on-screen list of example activities derived from common military exercises 2 as a guide to the user. This change improves the usability of the app without materially affecting its accuracy. Clothing Insulation and Vapor Permeability The insulation factor of a given clothing ensemble is designated I T (units are clo), whereas i m represents the vapor permeability of a clothing ensemble. 15 The evaporative capacity of an ensemble is represented by the i m /clo. 15 A parameter sweep was performed for clo values of 1.08 to 3.0 to capture a wide range of insulation types. The sweat rate estimate range was well above the minimum threshold of 125 ml/h. However, as end users are not expected to be able to provide clo or i m /clo values directly, the app has implemented an ensemble-based categorical input based on representative clothing examples. Common clothing ensembles were selected to provide adequate range of clo and i m /clo values without overwhelming the user with a long list of ensembles. Like metabolic rate, smaller changes in clo and i m /clo between categories had only a small impact on sweating rate, thus the final input provides three categorical selections. The three clothing ensembles selected were the Army Combat Uniform, Army Combat Uniform + Body Armor, and Mission Oriented Protective Posture Level 4 (i.e., protective uniform for chemical and biological threats worn in a closed configuration). Ambient temperature, relative humidity, and cloud cover were similarly analyzed and found to have significant influences on the output of the model. These were included as direct input variables. Default Values Based on the sensitivity analyses, and as part of the simplification of the app, we decided to use standard default parameters for certain variables. Body size was given default values of 79.8 kg and 1.8 m, on the basis of previous large surveys of Army soldiers documenting average values for height and weight. 16 The reason this assumption is required is not because height or weight independently influence sweating, but because the PW model requires that sweating rates be expressed per unit body surface area before 1143

3 converting to L/h and the range of expected body surface area differences create only small errors. Air velocity was standardized at 2 miles/hr (0.89 m/s) since this air velocity was used for the majority of experimentally observed sweating rates, and altitude was standardized at 30 feet (9.1 m) above sea level because of the marginal influence of higher altitudes on whole body sweating rate. 14 Independent Verification and Validation (IV&V) Once these simplifications had been completed, an IV&V was completed, comparing the SWET app to the original algorithm (Excel file). Test automation tools were created to execute over 10,000 input value combinations on both the original U.S. Army Research Institute of Environmental Medicine (USARIEM) algorithm and MIT Lincoln Laboratory app, and compare the results. Acceptability Testing We conducted preliminary user acceptability testing of SWET in collaboration with the U.S. Army Mountain Warfare School (AMWS) in Jericho, Vermont. We delivered questionnaires to AMWS via postal service, and instructors distributed questionnaires to students attending the school during July Several smartphones were provided to AMWS instructors and were used by 44 of their students. Following this, the students completed the questionnaires, which the AMWS instructors then returned to USARIEM. No personally identifiable information was collected. The protocol for distribution of questionnaires and collection of acceptability information from AMWS students was reviewed by the USARIEM Office of Research and Quality Compliance and was determined to be a nonresearch test protocol, in accordance with 32 Code of Federal Regulations (Department of Defense Policy for Protection of Human Research Subjects). Students were asked questions related to the ease of use of the app, and also related to the importance of hydration and drinking water planning in their experience. The questionnaire is included as an Appendix. All questions utilized a 5-point scale. Anchor words for all questions were: 2 = strongly disagree; 1 = disagree; 0 = neutral; 1 = agree; and 2 = strongly agree. Questionnaire results were analyzed using median, mean, and frequency distributions for each question. Subgroup Analysis The target of Nett Warrior app development is those soldiers whose Military Occupational Specialty/Area of Concentration (MOS/AOC) is more specifically boots on the ground infantry, artillery, etc. To obtain a more specific FIGURE 1. Screenshots of SWET main screen and (B) activity menu. 1144

4 set of responses from groups of soldiers who would be the most likely to use the app, we also conducted a subset analysis of responses from those groups whose demographics identified them as the individuals who would be more likely to use the app and for whom PM Nett Warrior is aiming its app development. Specifically, these were the boots on the ground MOS/AOC categories who had the potential to be unit leaders (rank of E-5 and above). RESULTS SWET App Sample screenshots are shown in Figure 1. The app provides the user with the amount of water required for the specified conditions in liters per hour. In addition, a Mission Planner tab adds the option to include the total number of people and the total time of the mission. The Mission Planner then calculates and reports total water needs in liters, 1-quart canteens, 2-quart canteens, and gallons. A sample screenshot of the Mission Planner is shown in Figure 2. IV&V The results of the IV&V demonstrated that the MIT-LL app is a faithful translation of the original USARIEM algorithm, with over 97% of the test cases showing exact accuracy. The remaining 3% showed a deviation of 0.01 or 0.02 liters/ hour, far below the published accuracy rate of ±0.125 liters/ hour of the USARIEM model. Acceptability Testing 44 individuals volunteered to complete questionnaires. All were male AMWS students with a range of ranks and MOS/ AOC categories (see Table I). 16 (36%) of the participants had been deployed. Results of acceptability testing indicated a high level of positive responsiveness to the app on the part of potential users. Mean and median responses ranged from 1 (agree) to 2 (strongly agree) for all ease of use and relevance questions, including SWET is easy to use, SWET is a better tool to predict my (or my unit s) water needs compared to any tool I currently have available ; and I would use SWET before a mission to predict my (or my unit s) water needs. Frequency distributions also showed a rightward skew in responses to all questions regarding ease of use and relevance, indicating a preponderance of agree and strongly agree responses. Table II shows mean and median values for all participant responses. Subgroup Analysis The MOS/AOC categories included in the subgroup analysis are starred (*) in Table I, and the mean and median values for their responses are shown in Table III (subgroup n = 15). In general, subgroup responses were similar to the overall group responses, which indicated largely positive TABLE I. Participant Demographic Information FIGURE 2. Screenshot of mission planner tab. MOS/AOC (Series Number) n % Infantry (11)* Student* 2 5 Signal (25) 2 5 Armor (19)* 5 11 Corps of Engineers (12) 3 7 Civil Affairs (38) 1 2 Field Artillery (13)* 2 5 Medical Career Management Field (68) 1 2 Transportation (88) 1 2 Special Forces (18)* 1 2 Psychological Operations (37) 1 2 Military Police (31)* 1 2 Rank CDT (none)* 2 5 PFC (E-3) 4 9 SPC (E-4) CPL (E-4) 1 2 SGT (E-5)* SSG (E-6)* 1 2 2LT (O-1)* 2 5 1LT (O-2)* 1 2 CPT (O-3)* 1 2 *Starred groups were included in the subgroup analysis (subgroup n = 15; Table III) for those individuals who are the targeted users of SWET in the field. 1145

5 TABLE II. Overall Acceptability of SWET: Mean and Median Responses, Whole Group Analysis Acceptability Question Mean Median n SWET is easy to use I understand what all the menu choices mean I understand what all the numbers mean in the menu The Mission Planner tool is easy to understand and use SWET provides information that is helpful to me in my MOS/AOC Being dehydrated is a major problem in my unit/mos/aoc We don t carry enough water with us on missions/ training exercises We carry too much water with us on missions/ training exercises We carry just the right amount of water with us on missions/ training exercises SWET is a better tool to predict my (or my unit s) water needs compared to any tool I currently have available I would use SWET before a mission to predict my (or my unit s) water needs The questions utilized a 5-point scale. Anchor words for all questions answered were: 2, strongly disagree; 1, disagree; 0, neutral; 1, agree; and 2, strongly agree. TABLE III. Acceptability of SWET: Mean and Median Responses, Subgroup Analysis a Question Mean Median n SWET is easy to use I understand what all the menu choices mean I understand what all the numbers mean in the menu The Mission Planner tool is easy to understand and use SWET provides information that is helpful to me in my MOS/AOC Being dehydrated is a major problem in my unit/mos/aoc We don t carry enough water with us on missions/ training exercises We carry too much water with us on missions/ training exercises We carry just the right amount of water with us on missions/ training exercises SWET is a better tool to predict my (or my unit s) water needs compared to any tool I currently have available I would use SWET before a mission to predict my (or my unit s) water needs The questions utilized a 5-point scale. Anchor words for all questions answered were: 2, strongly disagree; 1, disagree; 0, neutral; 1, agree; and 2, strongly agree. a Subgroup analysis includes potential unit leaders with a rank of E-5 and higher and a MOS/AOC of Infantry, Armor, Field Artillery, Special Forces, or Military Police. These are the starred (*) MOS categories in Table I. opinions. However, the subgroup tended to feel that too little water was carried. DISCUSSION The ability to rapidly and accurately plan Warfighters needs for drinking water in battlefield and training environments is a major unmet requirement for the Army and Department of Defense. The SWET app is a unique software-based smartphone/smart device app for prediction of water requirements under various conditions, based on complex models that incorporate metabolic rate, environmental parameters, and clothing insulative characteristics. 10,12 14,17 The app provides clear, easy-to-use screens for entering relevant mission parameters and displaying estimates of cohort fluid needs over specific periods of time. Military activities often require prolonged periods of work in hot environments while wearing body armor and other forms of protective clothing that augment heat storage and promote sweating. These situations can easily lead to moderate to severe dehydration if adequate fluid is not provided. When dehydration exceeds 2% of body weight loss ( 1.5 L), performance begins to decline and is worsened by environmental heat stress. 4 In addition to impaired performance, dehydration exacerbates morbidity associated with many other serious conditions including heat illness, orthostatic intolerance, diarrheal illness, altitude illness, and traumatic brain injury. 3,18 21 This then complicates clinical management of potential casualties who may suffer from one or more of these conditions. Overdrinking is another, less common but potentially fatal, problem associated with poor water planning and provisions. 6,8 Together, dehydration and hyponatremia account for more than 9,000 ambulatory hospital visits per year in the tri-service training environment. 22,23 Accurate, real-time estimates of water needs are key to preventing both overdrinking and under drinking. An important aspect of any new decision aid is that potential users find it to be user friendly and relevant to their daily life. Our goal in the present app development was that it should be geared toward unit leaders who conduct water planning for missions or training activities of varying lengths of time. Preliminary user acceptability testing indicated a high level of acceptability on the part of potential users of the SWET app. Frequency distribution, mean, and median values for participant responses indicated that the 1146

6 users tended to agree that the app is easy to use, helpful to a range of job categories, and the outputs are easy to understand. The subgroup of Nett-Warrior specific participants also showed responses that were similar to the overall group responses in terms of positive results for usability and applicability to their mission/mos/aoc (Table III). Limitations The goal of this manuscript was to describe the initial development and preliminary acceptability testing of a userfriendly mobile app for water planning in the military. In this context, we are cognizant of several limitations, which will require follow-up with the user community in the coming years. First, our acceptability testing was conducted on a relatively small sample (n = 44). Second, we do not have detailed information regarding the interaction of the users with the app and any potential errors that might have occurred in parameter entry or water needs calculations. Third, only a subset of the group had been deployed. Although we believe that individuals who have not previously been deployed can provide relevant feedback regarding potential usefulness in the field, this could be viewed as a limitation as well. Thus, although these individuals generally liked the app, and felt it was useful, they likely did not identify every aspect of the app, which might need updating or modification in subsequent versions. The preliminary acceptability testing did identify some areas for improvement, which will be the goal of subsequent versions. Finally, since SWET has been fielded (April 2015), we anticipate that we will receive additional feedback (and be able to make plans for modifications) once more potential users have had the opportunity to work with the app. Certainly, more formalized user acceptability testing with larger sample sizes will be helpful. In summary, the SWET will support a wide range of military users to include mission planners for combat and humanitarian missions, medical personnel, and unit leaders. SWET makes possible the seamless integration of paper doctrine and guidance 2 into a flexible user tool on the modern digital battlefield. Greater visibility and use of doctrine in this new format should improve education and aid in the prevention of fluid and electrolyte imbalance problems. Overall, responses from potential users were very positive with regard to ease of use and relevance to military training and deployment activities. Future versions of this decision-aid app will likely also have potential to the sport and wilderness medicine community and for disaster relief /humanitarian aid missions. ACKNOWLEDGMENTS We gratefully acknowledge Dr. Richard R. Gonzalez for his central work in developing the algorithm that was the basis for the development of SWET, and for his important input and suggestions in the development of the app. Development of the models to predict water loss and hydration requirements, as well as the transition of these models into specific smartphone app development, have been supported by U.S. Army Medical Research and Materiel Command Military Operational Medicine Research Program (Task Area 10, Hot Weather Operations and Hydration: Injury and Performance Optimization). REFERENCES 1. Force Development DirectorateU.S. Army Combined Arms Support Command: Potable water consumption planning factors. Available at Planning_Guide_rev_103008_dtd_Nov_08_(5-09).pdf; accessed September 21, Headquarters, Department of the Army and Air Force: Heat stress control and heat casualty management. Available at army.mil/assets/docs/publications/articles/2003/tbmed507.pdf; accessed September 21, Carter R 3rd, Cheuvront SN, Williams JO, et al: Epidemiology of hospitalizations and deaths from heat illness in soldiers. Med Sci Sports Exerc 2005; 37: Kenefick RW, Cheuvront SN, Palombo LJ, Ely BR, Sawka MN: Skin temperature modifies the impact of hypohydration on aerobic performance. J Appl Physiol (1985) 2010; 109: Cheuvront SN, Kenefick RW, Montain SJ, Sawka MN: Mechanisms of aerobic performance impairment with heat stress and dehydration. J Appl Physiol (1985) 2010; 109: O Brien KK, Montain SJ, Corr WP, Sawka MN, Knapik JJ, Craig SC: Hyponatremia associated with overhydration in U.S. Army trainees. Mil Med 2001; 166: Montain SJ, Sawka MN, Wenger CB: Hyponatremia associated with exercise: risk factors and pathogenesis. Exerc Sport Sci Rev 2001; 29: Kolka MA, Latzka WA, Montain SJ, Corr WP, O Brien KK, Sawka MN: Effectiveness of revised fluid replacement guidelines for military training in hot weather. Aviat Space Environ Med 2003; 74: Montain SJ, Latzka WA, Sawka MN: Fluid replacement recommendations for training in hot weather. Mil Med 1999; 164: Shapiro Y, Pandolf KB, Goldman RF: Predicting sweat loss response to exercise, environment and clothing. Eur J Appl Physiol Occup Physiol 1982; 48: Cheuvront SN, Montain SJ, Goodman DA, Blanchard L, Sawka MN: Evaluation of the limits to accurate sweat loss prediction during prolonged exercise. Eur J Appl Physiol 2007; 101: Gonzalez RR, Cheuvront SN, Ely BR, et al: Sweat rate prediction equations for outdoor exercise with transient solar radiation. J Appl Physiol (1985) 2012; 112: Gonzalez RR, Cheuvront SN, Montain SJ, et al: Expanded prediction equations of human sweat loss and water needs. J Appl Physiol (1985) 2009; 107: Gonzalez RR, Kenefick RW, Muza SR, Hamilton SW, Sawka MN: Sweat rate and prediction validation during high-altitude treks on Mount Kilimanjaro. J Appl Physiol (1985) 2013; 114: Gonzalez RR: Biophysics of heat transfer and clothing considerations. In: Human Performance Physiology and Environmental Medicine at Terrestrial Extremes. Edited by Pandolf KB, Sawka MN, Gonzalez RR. Natick, MA, Cooper Publishing Group, Bathalon GP, McGraw SM, Friedl KE, Sharp MA, Williamson DA, Young AJ: Rationale and evidence supporting changes to the Army Weight Control Program. Available at AD=ADA424879; accessed September 21, Pandolf KB, Stroschein LA, Drolet LL, Gonzalez RR, Sawka MN: Prediction modeling of physiological responses and human performance in the heat. Comput Biol Med 1986; 16: Carter R 3rd, Cheuvront SN, Vernieuw CR, Sawka MN: Hypohydration and prior heat stress exacerbates decreases in cerebral 1147

7 blood flow velocity during standing. J Appl Physiol (1985) 2006; 101: Brown JA, Riddle MS, Putnam SD, et al: Outcomes of diarrhea management in operations Iraqi Freedom and Enduring Freedom. Travel Med Infect Dis 2009; 7: Castellani JW, Muza SR, Cheuvront SN, et al: Effect of hypohydration and altitude exposure on aerobic exercise performance and acute mountain sickness. J Appl Physiol (1985) 2010; 109: Clifton GL, Miller ER, Choi SC, Levin HS. Fluid thresholds and outcome from severe brain injury. Crit Care Med 2002; 30: Armed Forces Health Surveillance Center: Hospitalizations among members of active components, U.S. Armed Forces, MSMR 2008; 15(3): Armed Forces Health Surveillance Center: Update: exertional hyponatremia, active component, U.S. Armed Forces, MSMR 2010; 17(3):

8 1149

9 1150