SYSTEM AND METHOD FOR MONITORING WELLBEING

Background of the invention

[0001 ] The present invention relates to a health/wellbeing monitoring system for providing real- time individualised risk-reduction feedback to a user.

[0002] The present invention also relates to a health/wellbeing monitoring system for providing real-time adjustment of an insurance policy premium.

Description of the Prior Art

[0003] Any reference herein to known prior art does not, unless the contrary indication appears, constitute an admission that such prior art is commonly known by those skilled in the art to which the invention relates, at the priority date of this application.

[0004] With the advent of wearable devices, individual users are able to monitor their individual health related data, such as heart rate, number of steps walked in a day, and other details of their personal activity.

[0005] This has the effect of motivating some individual users to try to achieve certain goals, such as walking a certain number of steps in a day, running a certain distance each day, or otherwise increasing their active time with other forms of exercise. This, of course, has consequential improvements in fitness, which correlates to improved overall wellbeing and reduced risk of sickness, with numerous benefits including an increase in anticipated average life expectancy.

[0006] When individuals seek an insurance policy, such as a life insurance or a medical insurance policy, an insurance company typically seeks to assess the individual’s risk profile prior to offering to insure the individual. When taking out medical insurance, it is typical for an insurance company to also exclude pre-existing medical conditions, at least for a certain period of time, from the insurance coverage. In other situations, such as when applying for life insurance, it is typical for individuals seeking the insurance to undergo a range of medical tests, such that the insurance company can calculate their risk associated with the individual and thereafter determine the premium payable for the insurance policy, using a pre-determined algorithm, based on factors such as age, family history, personal history of sickness, blood tests, urine tests, and other paramedical tests or questions. The premium for the life insurance policy is then typically set at the static point in time the policy is initially acquired and may be further adjusted according to a predetermined formula from year to year as the life insurance policy is renewed. Typically, when an insured individual reaches a certain age or has certain health risk factors, the insurance company may not be willing to insure an individual or may charge a higher premium (‘loading’), due to a high statistical risk they may have calculated.

Summary of the Invention

[0007]The present invention seeks to provide a health and wellbeing monitoring system for providing real-time individualised risk-reduction feedback to a user, such that the user may initiate preventive and / or remedial action to improve their fitness or other wellbeing criteria and reduce modifiable chronicdisease riskfactors.

[0008]The present invention also seeks to provide a wellbeing monitoring system for providing real- time adjustment of an insurance policy premium, based on action which an individual may undertake to seek to improve their fitness or other wellbeing criteria.

[0009] In one broad form, the present invention provides a wellbeing monitoring system for providing real-time individualised risk-reduction feedback, the feedback being provided to motivate a user to adjust a wellbeing profile of the user, and, to provide real time adjustment of a parameter of an insurance policy of the user based on any change to the user’s wellbeing profile.

[0010] Preferably, the system includes,

a plurality of wellbeing data input devices; and,

a processor, adapted to:

receive input data from said input devices;

process said input data, including by correlating said input data with predetermined data to produce remediation data; and, an output device, to provide feedback information based on said remediation data to a user, such that said user can thereby initiate action to improve their wellbeing profile.

[001 1 ] Also preferably, each said wellbeing data input device includes any one or combination of: a user wearable device, for input of user fitness/health data, including heart rate variability and/or ECG data, user movement data, such as steps taken or distance travelled, user sleep data, calories burnt, fall detection data;

a demographic/psychographic data input device, for input of age, gender, weight and/or height, profession and/or income, relationship status, home location, interests, travel frequency and/or destinations, and/or money spending habits, lifestyle data, employment status;

a psycho-social data input device, for input of social media data, personality assessment data and/or classification data;

a user self-reported data input device, for input of self-reported pain scores, journaling, self-entered medical data, nutritional habits, motivation levels, questionnaire responses;

a medical/ clinical data input device, for input of clinical data such as but not limited to, diabetes status or cholesterol levels, medical/electronic health records;

a health data input device, for input of calorie consumption data, blood glucose data, genetic/telomere data;

a financial data input device, for input of credit scores, payment and banking transaction history or applications, KYC data;

an environmental data input device, for input of location, location classification and annotation, pollution levels, crime rates; and,

a public or proprietary data sets input device, for input of population-level risk factors and statistics, longitudinal population-based outcomes data, insurance claims data, health outcomes, and/or survival data.

[0012] Also preferably, said predetermined data includes at least one of:

stratification model data; and,

predetermined user data.

[0013] Also preferably, said wellbeing data is at least partially processed in, or via an application installed on, a user device, such as, but not limited to a smart phone or watch.

[0014] Preferably, said wellbeing data is at least partially captured and/or transferred from said input devices to said processor via any one or combination of API, Bluetooth connectivity, Wi-Fi, Induction wireless, Ultra-wideband, ZigBee, Infrared wireless or any other packet and/or data transmission means.

[0015] Also preferably, said input data is processed to determine a wellbeing indicator based on the health risks of a user as calculated from the input data received from the wellbeing data input devices.

[0016] Preferably, said feedback information includes individualised health or fitness target goals, motivational factors, programs or the like, supplied via said output device to said user.

[0017] Also preferably, said output device includes an audio and/or visual output device.

[0018] Preferably, said system increases its accuracy with each input device added to the plurality of devices.

[0019] In a further broad form, the present invention provides a wellbeing monitoring system for providing real-time individualised risk-reduction feedback to a user.

[0020] In yeta further broad form, the present invention provides a system for adjusting the wellbeing profile of a user, comprising :

a plurality of wellbeing data input devices; and,

a processor, adapted to:

receive input data from said input devices;

process said input data, including by correlating said input data with predetermined data to produce remediation data; and,

an output device, to provide feedback information based on said

remediation data to a user, such that said user can thereby initiate action to improve their wellbeing profile.

[0021 ] Preferably, wherein each said wellbeing data input device includes any one or combination of:

a user wearable device, for input of user fitness/health data, including: heart rate variability and/or ECG data, user movement data, such as steps taken or distance travelled, user sleep data, calories burnt, fall-detection data; a demographic/psychographic data input device, for input of age, gender, weight and/or height, profession and/or income, relationship status, home location, interests, travel frequency and/or destinations, and/or money spending habits, lifestyle data, employment status;

a psycho-social data input device, for input of social media data, personality assessment data and/or classification data;

a user self-reported data input device, for input of self-reported pain scores, journaling, self- entered medical data, nutritional habits, motivation levels, questionnaire responses;

a medical/ clinical data input device, for input of clinical data such as, but not limited to, diabetes status or cholesterol levels, medical / electronic health records;

a health data input device, for input of calorie consumption data, blood glucose data, genetic/telomere data;

a financial data input device, for input of credit scores, payment and banking transaction history or applications, KYC data;

an environmental data input device, for input of location, location classification and annotation, pollution levels, crime rates; and,

a public or proprietary data set’s input device, for input of population-level risk factors and statistics, longitudinal population-based outcomes data, insurance claims data, health outcomes and/or survival data.

[0022] Preferably, said predetermined data includes at least one of: stratification model data; and predetermined user data.

[0023] Also preferably, said wellbeing data is at least partially processed in, or via an application installed on, user devices, such as, but not limited to a smart phone or watch.

[0024] Also preferably said wellbeing data is at least partially captured and/or transferred from said input devices to said processor via any one or combination of API, Bluetooth connectivity, Wi- Fi, Induction wireless, Ultra-wideband, ZigBee, Infrared wireless or any other packet and/or data transmission means. [0025] Preferably, said input data is processed to determine a wellbeing indicator based on the health risks of a user as calculated from the input data received from the wellbeing data input devices.

[0026] Also preferably, said feedback information includes individualised health or fitness target goals, motivational factors, programs or the like, supplied via said output device to said user.

[0027] Also preferably, said output device includes an audio and/or visual output device.

[0028] Also preferably, said system increases its accuracy with each input device added to the plurality of devices.

[0029] In a further broad form, the present invention provides a wellbeing monitoring system for providing real-time adjustment of at least one insurance policy parameter.

[0030] In a further broad form, the present invention provides a system for adjusting an insurance policy parameter, comprising:

a plurality of wellbeing data input devices; a processor, adapted to:

receive input data from said input devices; and, process said input data, including by correlating said input data with predetermined data, to produce correlation data; and, calculating a modified insurance policy parameter based on said correlation data.

[0031 ] Preferably, said insurance policy parameter includes any one or combination of: an insurance policy premium, price or pricing option, frequency of payments, level of cover, claim back values, number of providers that may offer cover, term of contract and optional extras and/or any other insurance policy parameter.

[0032] Also preferably, each said wellbeing data input device includes any one or combination of:

a user wearable device, for input of user heart rate variability and/or ECG data, user movement data, such as steps taken or distance travelled, user sleep data, calories burnt;

a demographic/psychographic data input device, for input of age, gender, weight and/or height, profession and/or income, relationship status, home location, interests, travel frequency and/or destinations, and/or money spending habits;

a psycho-social data input device, for input of social media data, personality assessment data and/or classification data; a user self-reported data input device, for input of self-reported pain scores, journaling, self-entered medical data, nutritional habits, motivation levels, questionnaire responses; a medical/ clinical data input device, for input of clinical data such as but not limited to, diabetes status or cholesterol levels;

a health data input device, for input of calorie consumption data, blood glucose data, genetic/telomere data;

a financial data input device, for input of credit scores, payment transaction history, KYC data;

an environmental data input device, for input of location, location classification and annotation, pollution levels, crime rates; and,

a public data sets input device, for input of population-level risk factors and statistics, longitudinal population-based outcomes data.

[0033] Also preferably, said predetermined data includes at least one of: stratification model data; and predetermined user data.

[0034] Preferably, said processor processes said data in real time according to input data received in real time from said wellbeing devices.

[0035] Preferably, said wellbeing data is at least partially processed in, or via an application installed on, a smart phone or watch.

[0036] Also preferably, wherein said wellbeing data is at least partially processed in a centralised data processor.

[0037] Preferably, said wellbeing data is at least partially captured and/or transferred from said input devices to said processor via any one or combination of API, Bluetooth connectivity, Wi-Fi, Induction wireless, Ultra-wideband, ZigBee, Infrared wireless or any other packet and/or data transmission means.

[0038] Preferably, wherein said input data is processed to determine a wellbeing indicator based on the fitness and health risks of a user as calculated from the input data received from the wellbeing data input devices. [0039] Also preferably, based on said correlation data, remediation information is produced and fed back to said user, in the form of targeted individualised interventions seeking to improve the well-being of the user.

[0040] Preferably, the well-being of the user is monitored for adherence of said user to said targeted individualised interventions and, said insurance policy parameters are adjusted according to user compliance.

[0041 ] Preferably, said system increases its accuracy with each said input device added to the plurality of devices.

[0042] In a further broad form, the present invention provides a method for setting an insurance policy premium or other parameter, comprising the steps of:

determining a premium or other parameter of an insurance policy;

monitoring wellbeing of a user by receiving input data from a plurality of wellbeing monitoring devices;

correlating said input data with stratification model data to produce remediation data; providing feedback information based on said remediation data to a user, such that said user can thereby initiate action to improve their wellbeing profile;

monitoring compliance of said user to said feedback information via said wellbeing monitoring devices; and,

adjusting said premium or other parameter of said insurance policy according to the user compliance and/or predicted user compliance.

[0043] In a further broad form, the present invention provides a method for

monitoring/adjusting the wellbeing of a user, comprising the steps of:

receiving input data from a plurality of wellbeing monitoring devices;

processing said input data, including correlating said input data with stratification model data to produce remediation data; and,

providing feedback information based on said remediation data to a user, such that said user can thereby initiate action to improve their wellbeing profile.

[0044] Preferably, in said correlating step: said input data is correlated with stratification model data and/or predetermined user data to produce said remediation data.

[0045] In a further broad form, the present invention provides a system for initiating an insurance adjustment of a user, comprising of:

at least one input device; a processor, adapted to:

receive input data from said at least one input device;

process said input data, including correlating said input data with predetermined data to produce remediation data; and

calculating an insurance adjustment based on said correlation data.

[0046] Preferably, said insurance adjustment includes at least one of:

adjusting one or more parameters of an existing insurance policy of a user; and, initiating a new insurance policy for a user.

[0047] In a further broad form, the present invention provides a method for initiating an insurance adjustment of a user, comprising the steps of:

monitoring a user, by receiving input data from at least one user monitoring device;

correlating said input data with predetermined data to produce remediation data; calculating an insurance adjustment for said user based on said correlation data; and, providing feedback information to said user, such that user can thereby initiate an insurance policy adjustment of the user.

[0048] Preferably, said insurance adjustment includes at least one of:

adjusting one or more parameters of an existing insurance policy of a user; and, initiating a new insurance policy for a user.

[0049] In a further broad term, the present invention provides, a method of determining an insurance offer to a user, comprising:

at least one input device;

a processor, adapted to:

receive input data from said at least one input device;

process said input data, including correlating said input data with predetermined data;

determine an output to: make an offer of insurance to a user;

make no offer of insurance to a user; or,

make a determination that there is insufficient data to either make an offer or make no offer, such that, said user can thereby provide further input data.

[0050] Preferably, a method as claimed in claim 31 , wherein said method is performed on a continuous or periodic real-time basis.

Brief Description of the Drawings

[0051 ] The present invention will become more fully understood from the following detailed description of preferred but non-limiting embodiments thereof, described in connection with the accompanying drawings wherein:

Fig. 1 illustrates a schematic view of a health/wellbeing system of the present invention;

Fig. 2 illustrates a schematic view of an alternative health/wellbeing system of the present invention, incorporating an insurance policy parameter adjustment mechanism;

Fig. 3 illustrates a schematic view of an exemplary embodiment of the present invention;

Fig. 4 illustrates some typical examples of external data sources which may be provided via input devices;

Fig 5 illustrates examples of self-entered data graphical user interfaces (GUIs);

Fig 6 illustrates a representation of a data aggregation and processing layer of a processor of the system of Fig 1 , 2 and 3;

Fig. 7 illustrates an example of an output device display which may be provided to a user;

Fig. 8 illustrates how various data sources may be integrated to provide an output to the user; Fig. 9 illustrates a further example of an output device display;

Fig. 10 illustrates a further example of an output device display;

Fig. 1 1 illustrates further example of output displays;

Fig. 12 illustrates an embodiment of risk reduction process;

Fig. 13 illustrates some risk influencing factors for single cause morbidities; Fig. 14 illustrates generic risk classification models;

Fig 15 illustrates various PRS score outlines;

Fig. 16 illustrates an exemplary process of an underwriting engine;

Fig. 17 illustrates will an example of a health risk and the output motivation information;

Fig. 18 illustrates a personalised health journey track;

Fig 19 illustrates another output display;

Fig 20 illustrates a schematic view of a system of the present invention showing various insurance decision options being provided to a user;

Fig 21 illustrates a continuous underwriting spectrum onto which a user may be assigned a position from 0 to 100, representative of a user’s health, wellbeing and / or other insurance-related risk and / or eligibility to receive an offer;

Fig 22 illustrates examples of information which may be presented to a user to initiate a user input response, so as to classify the user’s risk status, Fig 22(a) showing an example querying a user in relation to a potential risk of diabetes, and, Fig 22(b) querying a user in relation to a potential risk of a heart or blood condition;

Fig 23 illustrates an overview of a user’s journey within the system of the present invention;

Fig 24 illustrates further details of a user’s journey within the system of the present invention;

Fig 25 illustrates exemplary interactive information and data input and output between the user and the system of the present invention; and,

Fig 26, in Figs 26(a) and 26(b), illustrates the further interactive input and output between the user and the system of the present invention.

Detailed Description of Preferred Embodiments

[0052] Throughout the drawings, like numerals will be used to identify like features, except where expressly otherwise indicated.

[0053] Fig. 1 illustrates a schematic view of a wellbeing monitoring system in

accordance with the present invention which is adapted to monitor the wellbeing of a user. The wellbeing monitoring system, generally designated by the 1 , includes a plurality of input devices 2, a processor 3, and, an output device 4. [0054] Each input device 2 may take a variety of forms, each adapted to provide input data 7 indicative of a health or other wellbeing characteristics of a user 5. For example, an input device 2 may be a user wearable device, such as a wrist worn smart watch or a user input device, such as a smart phone or a third party device such as a smartphone application or public or proprietary data source.

[0055] A smart watch input device may typically include heart rate monitor, a means for measuring the number of steps taken by a user, the number of calories burned by a user, data pertaining to a user’s sleep patterns, etc.

[0056] Another typical input device 2 may be a smart phone device into which a user inputs information via a keypad or the like. The user may typically use such an input device to import their age, gender, weight and/or height, details of their profession and/or income, their relationship status, home location, interests, travel frequency, money spending habits, etc.

[0057] It should be appreciated by a person skilled in the art that a wide variety of input devices 2 may be utilised to import a wide variety of health, social, or other

characteristics of a user 5 into the system 1 , all of which may have some effect on a user’s well-being.

[0058] In Fig 1 , the processor 3 is adapted to receive the input data 6 from each of the pluralities of input devices 2, and then process this input data 7.

[0059] The processor 3 may incorporate a suitable memory device which stores predetermined data, 8, which may include stratification model data 8, that is, data which is relevant to the demographic group pertinent to the individual user, and/or,

predetermined user data, that is, data which is relevant to earlier collected data of the individual user.

[0060] The processor 3 is adapted to then correlate the user input data 7 with the predetermined data 8 to produce remediation data 9, that is, data which provides some form of‘rating’ of how the particular individual 5 compares with the general population, or at least a relevant demographic group(s) thereof, or a pre-determined benchmark rating of the individual user, for example, based on changed circumstances of the individual user. [0061 ] The system 1 shown in Fig 1 , also incorporates an output device 4, which provides feedback information, based on the remediation data 9, to the user 5, such that the user 5 can thereby initiate remedial action, based on this remediation data 9, to at least try to improve their personal health or wellbeing profile.

[0062] Throughout this specification, various terms are used, and these terms should be interpreted to have broad meaning. These terms include terms such as‘wellbeing’, ‘stratification’,‘predetermined user data’,‘remediation’,’insurance policy parameter’, etc.

[0063] For example, when the term‘wellbeing’ is used throughout this specification, this term should be understood to include not only the physical health markers of an individual, but also other factors which may include their mental or emotional state and any other factor which may affect their wellbeing, mortality, and/or morbidity and/or comorbidity risks.

[0064] The term‘stratification’ used throughout this specification should be interpreted to include a form of grouping of people based on a variety of factors, such as, but not limited to, age, gender, income, wealth, social status or any other factors.

[0065] The term‘predetermined user data’ used throughout this specification should be interpreted to include data pertaining to an individual user which has been provided by one or more input device and which may be stored in a memory of the processor, and which may be used for comparison with data obtained at a later time, typically indicative of a change in the current or future circumstances of a user. This may, for example include an indication from a user’s social media that they have had or are going to have children, that they have been in an accident, that they have upcoming travel plans, or any other changed circumstance, lifestyle, etc. of a user.

[0066]The term‘predetermined data’ used throughout this specification should be interpreted to include‘stratification data’ and/or‘predetermined user data’.

[0067] The term‘remediation’ used throughout the specification should be interpreted to include any factors which may improve or correct the health of an individual.

[0068] The term‘insurance policy parameter’ used throughout this specification may include any or all parameters that may have an effect on or make up a user’s insurance policy. These may include costs, frequency of payments, level of cover, claim back values, number of providers that may offer cover, term of contract and optional extras.

[0069] The present invention therefore provides a system which, in effect, correlates user-specific input data with predetermined data, including generalised stratification model data or individual predetermined data, to produce remediation data output which is indicative of goals which a user should seekto achieve to improve their wellbeing profile.

[0070] It will be appreciated that each wellbeing data input device 2 may take a wide variety of forms. Preferred, but non-limiting, examples of various input devices 2 include hand held devices capable of user input such as smart phone applications, third party data sources (Fatsecret, 23andMe, bank feeds, MyFitnessPal, iHealth or Experian, for example), consumer health aggregation services (Apple HealthKit or Google Fit), wearable health devices (FitBit, Apple Watch, Garmin devices, etc. ), clinical data (My Health Record, Cerner, Epic EMRs and / or EFIRs), public or proprietary data sets (e.g. claims history data) and social media application and productivity tools (Facebook, Twitter, email service providers, Microsoft Office or Linkedln) input devices.

[0071 ] The processor 3 may be embodied in a variety of ways. In certain embodiments, the processor 3 may, for example, be at least partially embodied on a user’s smart phone or watch, or within other medical or clinical data devices which have appropriate processing circuitry. In other embodiments, the processor 3 may be at least partially embodied in the form of a remotely located processor.

[0072] The data 7 may be transferred from each input device 2 to the processor 3 in a variety of forms. For example, the data transfer may be via any one or combination of API, Bluetooth connectivity, Wi-Fi, induction wireless, ultra-wideband, ZigBee, infrared wireless or any other packet and/or data transmission means.

[0073] The output device 4 may also be embodied in a variety of forms, for example, the output device 4 may be in the form of a visual display unit such as a screen of a smart phone, watch or any other output device, an audio output device such as a speaker, and/or a combined audio/visual device. [0074] In use, a user 5 will typically be able to see and/or hear some form of display of remediation data 9 via one or more output device 4, such that they can thereby take appropriate action, if desired, to improve their fitness or well-being.

[0075] The output device 4 may therefore typically provide individualised health or fitness target goals, motivational factors, programs, digital signposting to third party services or the like, to the user 5.

[0076] As will be appreciated by people skilled in the art, adding additional input devices 2 will naturally improve the accuracy of the overall system 1.

[0077] As will also be appreciated, as the user 5 receives a display of remediation data 9 on the output device 4 and initiates appropriate action, such as additional exercise, or the like, to achieve the appropriate goals which are provided on the output device 4, then the health/wellbeing of the user 5 is thereby improved and the insurance policy parameters may be likewise adjusted.

[0078] Fig. 2 illustrates a schematic view of a variation to the wellbeing monitoring system which may typically be utilised by an insurance company to adjust an insurance policy parameter based on the altered health/wellbeing of the user.

[0079] As per the device of Fig. 1 , the system for adjusting an insurance policy parameter, as illustrated in Fig. 2 and generally designated by the 21 , includes a plurality of input devices 22, a processor 23, and, an output device 24.

[0080] In addition to the processor receiving input data 26 from each of the input devices 22 and, correlating this composite input data 27 with predetermined data 28 to produce remediation data 29, the processor is further adapted to calculate a modified insurance policy parameter based on the remediation data of the user 25.

[0081 ] A conventional insurance policy parameter 30, may thereby be adjusted, by an adjustment factor 32, generated by the processor 23, and, based on the remediation data 29, produce an adjusted insurance policy parameter 31.

[0082] The insurance policy parameter 30 may be any typical parameter of an insurance policy such as, but not limited to, the price of the insurance policy over a periodic, for example a yearly basis - or pro-rated amount over a shorter period, such as a month, week, or day, another pricing option, the frequency of payment of the premium, the level of cover, claim back values, the term of the insurance, and/or any other parameter or combination of parameters of an insurance policy.

[0083] That is, in addition to the user 25 being provided with remediation data from the output device 24, the user 25 may be further encouraged or incentivised to improve their health/wellbeing by receiving an adjustment of their insurance policy parameter. For example, an adjustment, such as a reduction or increase in the price of their annual premium - or monthly, weekly, or daily pro-rated equivalent thereof - of their insurance policy, may incentivise a user to improve their individual health/wellbeing.

[0084] The user 25 is thus provided with targeted individualised information to

incentivise them to improve their health or well-being by their compliance with the information provided, such as, achieving certain health goals.

[0085] The system 21 , as shown in Fig. 2, monitors the user's adherence to the targeted individualised information via the input devices 22. That is, the input devices 22 directly monitor the user's 25 activity or other health data to confirm compliance with the goals or other information provided to them via the output device 24, to thereby ensure that they are achieving their targeted goals or exercise regime.

[0086] The system can take a user's predetermined data, profile, and various data input sources to provide personalised, needs-based insurance cover options. More specifically, the system is able to calculate from a user's profile and various input data (including, but not limited to, a user's banking and financial transaction data, mortgage loan information, health profile data, claims history, social media sources, and any other relevant external or self-reported data source), personalised insurance policy cover parameters based on likely protection needs (e.g. tailored cover product configuration and pricing based on the fact that a user owns a $1 M property with an outstanding mortgage loan balance of $450,000 payable over the next 22 years and is currently spending $6,439 per month on living expenses), the system can dynamically and in real time or near-real-time determine the most appropriate level of life insurance cover by way of payout needs to meet those mortgage payments over a predetermined period.

[0087] The dynamic nature of this needs-based cover means that, by way of example, as mortgage payments reduce the outstanding balance of the loan or as new debt may be taken on or new life events occur (e.g. the birth of a child as determined based on ingested social media data) or as a user's health profile changes (e.g. as a user's Type 2 diabetes risk increases over time or as hypertension risk reduces over time), the payout and cover and policy options and benefits may change accordingly. The requisite payout for a term life policy, for example, may decrease as a user's mortgage reduces over time; equally, the user may be proffered additional cover to cater for the birth of a child (as triggered within the system by birth records or social media updates) or the increased likelihood of medical expenses as a result of changing health

circumstances or adherence to the system-recommended health interventions or a policy's required premium payments may be automatically paused because of the system learning of a user's recent unemployment status (e.g. via Linkedln or a change in deposits in the user's bank account).

[0088] The present invention therefore provides a system and method which can initiate a personalised new insurance policy, or, amend a current user’s insurance policy based on user need, in real time, or close to real time. Such policy parameters are not limited to life insurance, but may apply more broadly to any form of protection cover, e.g. D&O insurance, travel insurance, etc.

[0089] Such a system and method includes at least one input device which monitors a user's data. For instance, this could be the monitoring of the user’s Facebook or other social media account. By way of example, monitoring the social media account of a user may provide input data to the system of the present invention which, for example, indicates that the user may be getting married, having a child, going on a holiday, etc.

As will be appreciated by a person skilled in the art, such a circumstance would typically warrant a user's insurance premium being adjusted, or, a user perhaps taking out a new insurance policy.

[0090] By such monitoring of the user’s social media account for receiving such appropriate data from some other form of input device, the processor of the system of the present invention, upon receipt of such data, correlates this new information with previous information to produce appropriate remediation data, and thereby calculate either an adjustment to the current user’s insurance policy based on these changed circumstances, or, provide a user with a recommendation that they should initiate a new form of insurance policy. For example, if the user is having a child, then the level of a user's life insurance might typically be increased. Likewise, if the user is indicating on the social media account that they are about to embark on a skiing holiday, then the user may be prompted to initiate a new travel insurance policy to cover this type of travel. As will be appreciated, the system and method of the present invention therefore provides an automated system in real- time or close to real-time to either automatically effect an adjustment to an existing insurance policy premium, or, the initiation of a new insurance policy, etc.

[0091 ] Whilst the invention has been hereinbefore more generally described with reference to Fig. 1 and 2, some more specific details, implementations and exemplary embodiments will now be hereinafter described with a view to illustrating the details, variations and modifications of the invention.

[0092] Fig 3 illustrates a schematic view of an exemplary embodiment of components of the invention. The system 40, shown in Fig 3, incorporates a data aggregation and processing layer 41 , a risk - stratification engine 42, a precision predictive life

expectancy engine (2PLE) 43, and, a personalised underwriting engine 44. The system 40 consequently provides an output to a preventative risk reduction and remediation intervention platform 45, including personalised health tracks, and also outputs price, cover options 46, policy cover recommendations 47, etc. Each of these features of the exemplary embodiment will now be described.

[0093] The data aggregation and processing layer 41 , enables ingestion, cleansing, normalisation, pre- and post-processing, and secure storage of both manually entered (‘active’) and automatically collected (‘passive’) data from a variety of sources 48. This data may be curated for further processing.

[0094] The risk-stratification engine 42 may be a continuous risk-classification and scoring engine, which applies a variety of stratification models to output a range of risks, including - but not limited to - Years Lived with Disability (YLD), Quality-Adjusted Life Year (QALY), personalised morbidity and comorbidity risks, Quality of Life (QoL) score, insurance claims likelihood, and, hospitalisation risk over any given projected timeframe.

[0095] The precision predictive life expectancy engine (2PLE) 43, is an engine which processes data ingested from its data aggregation and processor layer 41 , The outputs of the risk stratification engine 42 typically may include a precision predictive life expectancy graph for an individual, factoring in a broad range of inputs and behavioural characteristics, including, but not limited to, biometric data, environmental data, psychographic data, current and past medical history and existing conditions, and lifestyle behaviours.

[0096] The personalised underwriting engine 44, may output continuous life and health and other insurance policy pricing, based on outputs from the data aggregation and processing layer 41 , the risk stratification engine 42 and the precision predictive life expectancy engine 43. It may also provide recommendations of policy cover options based on the individual’s personal profile and other data attributes known, derived and/or inferred through the data aggregation and processing layer 41 , the risk stratification engine 42 and the precision predictive life expectancy engine 43. These outputs 46 and 47 may be either standalone or may be subsequently transmitted over a network to be ingested by a third party’s policy administration or quoting system, for example, that of an insurer or reinsurer.

[0097] The preventive risk-reduction and remediation intervention platform 45, may include personalised health tracks. This may include dynamic digital platform

information, third party signposting, output to a third party device, such as smart speakers or fitness tracker or other wearable device, third party application such as via an API or batch transfer, and GUIs, encompassing targeted, precision interventions for the user based on the outputs of 42, 43 and 44, to provide motivation, tailored content, personalised health‘tracks’ and, goals. A continuous feedback loop facilitates ongoing re-rating of a user’s risk profile per 42 and 43 and the output thereby provides continuous underwriting of a user’s insurance policy and/or premium.

[0098] Any underwritten pricing and cover output 46, and/or policy cover recommendations for sell/upsell/cross-sell 47 may also be output, as hereinbefore described.

[0099] The data aggregation & processing layer, 41 as hereinbefore described, performs the following functions: via connection to a variety of data sources, this layer 41 extracts, ingests, and transforms big and disparate data, including structured, semi- structured, and unstructured data, from its source of origin into a system where it can be stored and analysed. [0100] Data can be streamed in real-time, via manual entry by a user, for example, via the risk-reduction services 45, or, ingested in batches, subject to technical constraints and requirements, for example, latency, third party restrictions on data scraping via API, etc.

[0101 ] Figs. 4 and 5 illustrate some examples of data inputs 50 to the system 51. Data sources are intentionally unrestricted and may include both health and non-health data attributes, including, but not limited to, demographic/ psychographic data (e.g. age, gender, height, weight, interests, lifestyle attributes), psycho-social data (e.g. social media data, personality assessment and classification data), wearables and fitness data and biomarkers (e.g. ECG, HRV, sleep recording data), medical/clinical data (e.g.

EHR/EMR, diabetes status, cholesterol levels), user self-reported data, encompassing qualitative, quantitative, health, and non-health attributes, e.g. self-reported pain scores, journaling, self-entered medical data or questionnaire responses), third party health data, e.g. pre-programmed HbA1 C or blood glucose data synced directly into the ingestion layer via API or other data passively ingested via API or Bluetooth or over any similar network/transport protocol (e.g. calorie consumption data synced via

MyFitnessPal or other such similar service), financial data (e.g. credit scores, payment transaction history, KYC data), environmental data (e.g. via GPS, location, and location classification and annotation, pollution levels, crime rates, etc.), public data sets (e.g. population-level risk factors and statistics, longitudinal population-based outcomes data), other physiological/biometric data: heart/biological age outputs, health and lifestyle behaviour data, (e.g. nutritional habits, motivation levels and genetic, epigenetic, and telomere data) and self- reported data (family medical history details, a user’s motivation to change, etc.).

[0102] Fig. 6 illustrates how, at the data collection level 62, different types of data may be brought together, organised and arranged in such suitable configuration as to enable further refining of the data and arrangement of it to support any requisite enhancement, cleansing, and/or improving of the raw data.

[0103] Any such data may be captured by the data collection and ingestion layer via API, Bluetooth connectivity or any other similar secure means of packet and data transmission and held securely in a data store, e.g. Amazon S3 or Microsoft Azure or in some other form of on premise or cloud database or other storage mechanism, 67. [0104] Data may be stored in either a standard relational database 67 or, more likely, given the volume, type, and velocity of the incoming data, in some form of big data storage tool, such as HDFS (Hadoop Distributed File System), GlusterFS or Amazon Simple Storage Service (Amazon S3). The choice of storage solution may be a function of implementation-specific requirements and constraints, e.g. requirements for the provision of elasticity in storage and performance without affecting active operations, scalability, nature of data, and whether large, distributed storage solutions are required.

[0105] In the processing layer 66, data is routed to the required destination(s). This may be performed via batch process, e.g. via Apache Sqoop, near-real time, or real-time processing, subject to requirements and operating constraints.

[0106] Any required analytic processing which cannot be performed in layer 64 may be carried out in the data query layer 65. Post-storage and processing of the data and any required ETL, data summarization, ad-hoc query, and analysis of the relevant dataset(s) can be undertaken in the data query layer. This may include, for example, big data analysis of trends and underlying correlations for the purpose of rendering output to users at the presentation layer 63 to show improvement in user health profiles or risk ratings or comparative distributions, for example, how a particular individual compares to others similar individuals. Equally, other large-scale data analysis which cannot be performed at the risk-stratification or predictive life expectancy layer 64 may be undertaken here.

[0107] Security 68 is preferably implemented at all layers and spans across ingestion, storage, processing, visualisation, etc. Security controls may vary depending on the specific implementation, but may include:

- Big Data Authentication: for example, the Kerberos protocol as a mechanism for user authentication;

- Access Control: restricting users and services to only those data which they are permissioned for and need to have access to;

- Encryption and Data Masking: to secure access to sensitive data, both in

transit and at rest;

- Auditing data and other security controls: e.g. via log files and access

attempts, as well as adherence to privacy requirements (e.g. GDPR) and information security best practice (e.g. ISO 27001 :2015, SOC 2). [0108] Data quality is preferably managed via ongoing monitoring, auditing, testing, and controlling of the data 69. Continuous monitoring of data is an important part of the governance mechanisms and may include:

- Data Profiling and lineage: These are the techniques to identify the quality of data and the lifecycle of the data through various phases. In these systems, it is important to capture the metadata at every layer of the stack so it can be used for verification and profiling;

- Data Quality: meeting its needs and satisfying the intended use;

- Data Cleansing: various solutions to correct any incorrect or corrupt data;

- Data Loss and Prevention: Policies and procedures for mitigating and

remediating data loss. Identification of such data loss needs careful monitoring and quality assessment processes.

[0109] Referring back to Fig. 3, the risk-stratification engine 42 performs a number of functions, namely, based on a spectrum of data inputs, varying from minimal, e.g. age, gender, country through to extensive longitudinal data encompassing a user’s medical history and genetic data, psychosocial data, etc. The risk classification engine is preferably able to determine and display as output to the user or designated third party, e.g. insurer, single-cause or multi-cause morbidity and / or mortality risk factors on an individualised basis over a defined time horizon and with a known degree of confidence or certainty in the prediction.

[01 10] Fig. 7 illustrates an example of an output display 70 including the ability for a user to validate the internal model by entering additional information, if requested, and/or entering in or providing permission to collect further data points relating to the individual which will further improve the accuracy of the prediction, e.g. by entering an hbA1 c level or connecting a continuous blood glucose monitoring device to the solution.

[01 1 1 ] Fig. 8 illustrates how the system leverages the data ingested by the data aggregation and processing layer 41 of Fig. 3, and, calculates and displays or provides as output to the user or designated third party, e.g. insurer or reinsurer, an overall risk score, which may be a sum of co- morbidity risks plus single-cause risk factors to produce a score, nominally out of 100. This may be further supplemented by a Quality of Life (QoL) score, again, nominally out of 100, which is the sum of weighted, aggregated values comprising key‘performance indicators’ of what would determine a healthy life, for example: - Overall health fitness levels.

- Income level.

- Psycho-social and environmental status.

Specific risk and illness likelihood, severity, comorbidity-adjusted status (e.g. Type 2 diabetes may be weighted more heavily than Hodgkin lymphoma because of a formula of: attendant co-morbidity weighting ^* incidence likelihood ^* mitigating actions (such as lifestyle factors or lack of genetic predisposition).

[01 12] With such QoL indicators known and weighted coefficients also known, this QoL score can thus be calculated:

Where, a1 is the specific indicator and b1 is the weighted coefficient of the indicator

(a1 ).

[01 13] Figs. 9 and 10 exemplify how single-factor results may be presented to the user based on a variety of models which will output aggregate assessments at the individualised level of a number of human disease burden impacts, for example:

- Disability adjusted life year (DALY);

- Years of Life Lost (YLL);

- Years of Life Lived with Disability (YLD);

- Quality Adjusted Life Years (QALY);

- Specific morbidity risk (e.g. risk of developing pre-diabetes, Type 2 diabetes, etc.);

- Mortality risk.

[01 14] These will also take inputs from the 2PLE engine 43, shown in Fig. 3.

[01 15] Fig. 1 1 exemplifies how, rather than calculating these measures at a population level or broad cohort level, however, the system may determine personalised scores and calculations based on the broad range of data ingested at 41 (see Fig. 3), assumptions and validated data which exists already in the system, single- and all cause morbidity and mortality factors, and an individual’s longitudinal data to measure and re-assess frequently over time (for example, every hour or day). [01 16] Furthermore, the risk classification and segmentation engine may aggregate the aforementioned with extant or configured, e.g. by an insurer or reinsurer, claims data to calculate and output claims likelihoods across any life or health risk class (e.g. income protection, TPD, short-term disability, accident and health, term life insurance, etc.), as well as hospitalisation likelihood, whether this is claimed or not under a policy, and further enables the selection of a relevant time horizon for the prediction (e.g. greater than or equal to 1 year), such as shown in Fig. 11 (a).

[01 17] In conjunction with cost profile data, ingested data from external or internal sources, such as banking feeds and transaction data or credit information, and weightings, configurable by an insurer or reinsurer, this can further output an

individualised likely cost profile, with associated confidence level, for a given user of the solution, such as shown in Fig. 1 1 (b). The system can also further output a target insurance policy/premium offer in the form of a currency amount across any relevant product line (e.g. health, TPD, IP, etc.), such as illustrated in Fig. 11 (c).

[01 18] Fig. 12 illustrates, by way of example, how the system may further disaggregate single-cause and all-cause morbidity, hospitalisation, claims, and other risk factors and configurable weightings and outputs to the user personalised goals and targets which, if adhered to, will result in a quantifiable risk-reduction (which may likewise correspond to an adjustment in insurance policy parameter), based on various known risk attributes for a given user.

[01 19] Fig. 13 and 14 show how, in addition to individualised risk scores, the system may provide risk categories based on the risk-stratification engine’s all-cause algorithms and the user profile and other data which can be used for more generalised

interventions within the solution, or, that of a third party, e.g. insurer, as a basic risk assessment mechanism which can likewise be extrapolated to others who share similar characteristics in the event a third party does not have its own user data at a sufficient level of granularity (e.g.“those who share your characteristics typically have a MEDIUM risk profile”, etc.)

[0120] It should be noted that the above stratifications can be achieved with very little requisite data, e.g. age, country, gender, albeit with less degree of certainty than when a greater depth, and/or breadth of data is provided. [0121 ] Over time, the risk models themselves become trained and validated via a classification of quality and type of data inputs provided by third party data sources or by the user. For example, while the risk stratification engine may assign a confidence score of 56% to a given risk factor, the confidence score may increase in such fashion as:

- User integrates a third party app which records their hbA1 c and continuous blood sugar levels, to provide, for example, a 28% improvement in risk factor confidence;

- User opts to connect his/her EHR which contains a clinically-recorded hbA1 c record from 42 months prior, to provide, for example, a 12% improvement in risk factor confidence, and/or;

User self-enters his/her hbA1 c from an at-home test, to provide, for example, a 6% improvement in risk factor confidence.

[0122] This self-learning and reinforcement model facilitates artificial neural networks (ANN) which are used to identify values based on training/validation data and lead to automatic classification of new data in the system. ML methods such as K-means clustering, SVM, Case-Based Reasoning and others may be employed to train and improve the risk stratification engine.

[0123] Referring back to Fig. 3, the precision predictive life expectancy engine (2PLE)

43 of the multi- factor personalised and predictive life expectancy model may aggregate all of the outputs of the risk stratification engine, shown in Fig. 6, and ongoing data ingestion, to output a personalised, precision life expectancy for each individual user.

[0124] Following querying, processing, and analysis of all ingested data and filtering thereof through the risk classification engine, an individual score may be assigned to each risk vector and Risk- Influencing Factor (RIF) associated with a user’s risk profile.

[0125] This may be combined with other quality and quantity of life factors, such as YLL, DALY, and an overall QoL score and further multiplied by the relevant comorbidity factor according to the requisite debiting of any user risk score as a result of inferred co morbidity penalties. This is offset by any mitigating factors the user may have, whether in the form of contra-indications by virtue of lack of strong genetic predisposition, lower risk biomarkers (Vo2 max, RHR, blood sugar levels, etc.), positive health behaviours such as nutritional habits and weight management, progress towards and/or achievement of system-generated goals, and psycho- social and other influencing factors (e.g. strong social network and graph, environmental factors such as proximity to open spaces, and living in a low-crime neighbourhood, etc.)

[0126] Fig. 15 shows that the output of this as a Predictive Risk Score (PRS). This may be an aggregate weighted summation of all the above risk factors and an associated confidence level based on availability of data and self-diagnosed system confidence in the accuracy of its prediction.

[0127] This 2PLE output provides the user or insurer with a constantly updated, real time life expectancy score over any given timeframe, based on multi-factor analyses including, but not limited to: demographic/psychographic data (age, gender, height, weight), psycho-social data (e.g. social media data, personality assessment and classification data), wearables and fitness data and biomarkers (e.g. ECG, HRV, sleep recording data), medical/clinical data (e.g. EHR/EMR, diabetes status, cholesterol levels), user self-reported data, encompassing qualitative, quantitative, health, and non health attributes, e.g. self-reported pain scores, journaling, self- entered medical data or questionnaire responses), third party health data, e.g. pre-programmed HbA1 C or blood glucose data synced directly into the ingestion layer via API or other data passively ingested via API or Bluetooth or over any similar network/transport protocol (e.g. calorie consumption data synced via MyFitnessPal or other such similar service), financial data (e.g. credit scores, payment transaction history, KYC data), environmental data (e.g. via GPS, location, and location classification and annotation, pollution levels, crime rates, etc.), Public data sets (e.g. population-level risk factors and statistics, longitudinal population-based outcomes data), other physiological/biometric data: Heart/biological age outputs, Health and lifestyle behaviour data, e.g. nutritional habits, motivation levels , genetic, epigenetic, and telomere data.

[0128] While current health-adjusted life expectancy models are based on one or more life tables and then combined with a prevalence-based measure of disability or of health-related quality of life based on limited population level data, the 2PLE model redefines this life expectancy expression by not only adjusting the number of life years projected to be lived by the output of the risk stratification model and single-cause and all-cause factors described above (i.e. an overall health score based on a far broader depth and breadth of data points and risk classification instruments), but also constantly adjusts the individual’s 2PLE score based on a feedback loop based on real-time or near-real time ingestion of data from the user (such that, by way of example, the recording of acute lower back pain for a continuous period of 4 days may lead to an increase in the underlying MSK risk factors which may in turn‘debit’ the user’s life expectancy by, for example 1 % per year, leading to the following adjustment to total life expectancy over a time horizon of, say, 45 years):

(0.01 * 365) * 45 = 164.25 days

[0129] Predictive personalised life expectancy can thus be expressed as follows:

where:

- 2PLE is health and risk-adjusted predictive life expectancy;

- x is the exact age for which life expectancy or health adjusted-life expectancy is to be estimated for an individual;

- i is an index representing the lower limit (x) of the age interval (x, x + 1 );

- Li is the number of life-years lived in the age group (x, x + 1 );

- lx is the number of survivors at age x;

- PRSi is a score or weight representing the Individualised risk for the age

group (x, x + 1 ), with PRSi = 1 indicating highest possible risk of mortality within the time period and PRSi = 0 indicating zero risk of mortality within the timeframe;

- Ci is the confidence in the PRSi, expressed as a value between 0 and 1 , where 1 is certain and 0 is no level of confidence, and n is the total number of age groups in the life table.

[0130] Fig. 16 shows, how, as a result of the data captured and ingested, the system may facilitate a number of advantages, including a streamlined, continuous underwriting process delivered via an app or similar technology, which reduces the typical underwriting time from weeks and days to minutes based on user data ingested over time and the 2PLE and risk stratification components as outlined above. This enables automated underwriting of an individual in most - if not all cases - without recourse to medical or paramedical examination, additional medical questionnaires, or blood or urine sample. [0131 ] A further advantage is the dynamic, real-time or near real-time or batch continuous underwriting pre- (‘prospective underwriting’) or at-the-point-of-purchase underwriting based user interaction with the solution, which updates or changes to data and a user’s risk profile or predicted life expectancy over time (e.g. as a result of lead indicators and predictors gleaned by the solution which alter the user’s risk profile or based on the user achieving system- recommended goals for improvement or maintenance of health), i.e. continuous, dynamic assessment of risk on single-cause and all-cause morbidity, mortality, and YLL factors, etc.

[0132] This may be further augmented by enabling the insurance policy parameters to be determined based on a system projection of a user’s likelihood to achieve certain risk-reduction goals over a pre-determined period. For example, rather than simply pricing a user’s premium based on their current, point-in-time, static risk profile, the policy price may be set at the likely level of risk the system anticipates the user can achieve over, say, 3 years, based on similar users’ risk- reduction efforts and the system’s continuous training and learning overtime.

[0133] A further advantage is the generation of individualised premium pricing and cover which may be based on pre-configured defaults, e.g. based on current meet-the-market pricing criteria or online broker data of product and pricing mix in the user’s given geographical jurisdiction, products available in the market or other criteria.

[0134] Pricing may either be determined by the solution itself or the user’s risk stratification outcome data, health score, and 2PLE can be passed through to an insurer or broker or reinsurer or other such insurance manufacturer or distributor in order to price based on their appetite e.g. based on insurers’ own or third party costs and weightings for lifetime costs.

[0135] Alternatively pricing may be determined based on typical projected lifetime costs which may be borne to service a given customer based on their projected claims profile.

[0136] In addition to pricing and products available to a given customer, the system may also recommend new products or policy considerations 47 based on be behavioural and risk analysis of the user’s data or on ingested data sources, including - but not limited to - life event triggers from social media or changes in financial data via bank feeds, which can be used for the purpose of real-time or non-real-time upsell purposes, e.g. with geofencing a user’s location and identifying it as being in the vicinity of an international airline terminal, travel insurance can be dynamically priced and proffered to the user, or, based on a user’s specific life stage and changes to their situation, e.g. based on discerning the user is in the process of purchasing a family home or has recently accepted a new job at a higher salary than previously or has paid off a debt, can be used to trigger specific policy cover considerations and upsell/deaccumulation options.

[0137] This underwriting may be undertaken on all users, not just those who are existing policyholders, therefore enabling an insurer and a prospective customer to understand future prospective risk, as well as point in time risk.

[0138] The underwriting engine can be further augmented via the upload or

transmission to it, e.g. via API or batch file transfer or similar, of various life tables or claims history data in bulk in order to feed into the underwriting engine’s core Machine Learning (ML) algorithms to effect a more custom and nuanced pricing and cover mix, subject to the needs of a given insurer.

[0139] The risk and underwriting outcomes may be mapped to standard/custom underwriting types, for example‘preferred’,‘standard’, etc.

[0140] The solution further supports a dynamic feedback loop between the risk-rating engine, 2PLE and data components to support highly targeted, individualised

interventions based on a user’s unique health, psycho-social and other profiles and risks.

[0141 ] Based on any given user’s specific disaggregated and aggregate risk factors (e.g. risk of Type 2 diabetes and risk of comorbidities associated with diabetes and/or overall life expectancy risk factors and quality of life score), the system may dynamically determine one or more interventions.

[0142] These may include in-app interventions: information, advice, guidance, and features designed to raise awareness of risk factors and reduce risk and/or encourage adherence to a health plan to reduce risk.

[0143] This may also include smart signposting, based on user data such as

geographical location, socio-economic status, past medical history, etc., facilitate joined- up, frictionless healthcare coordination by digitally signposting users to specific primary/allied healthcare providers or specialists (e.g. diabetes counsellors).

[0144] This may also include engagement and adherence tools, leveraging behavioural psychology and/or nudge/boost theories/motivational interviewing to further classify users based on propensity and willingness to change and likelihood to achieve desired outcomes.

[0145] Fig. 17 illustrates, by way of example, that, for each user with a risk classification, the output may subsequently be used for targeting any given specific health intervention (e.g. T2D).

[0146] Furthermore, through a self-learning model, the system is able to generate for the user one or more‘health journeys’ comprising the effect-adjusted joining-together in a sequence of relevant targeted programs, interventions, communications, etc., which enables loosely coupled modules to be joined together as components in an organised program. For example, given any number of possible module interventions within the system, the system may, for example, prioritise 5 in a sequence for the user based on, for example:

- User’s level and type of risk;

- Urgency (i.e. based on time horizon of risk);

- Motivational factors and psycho-graphic data (individual drivers of change, what has worked previously for the user);

- Past history (e.g. pre-existing medical may contra-indicate a given

intervention); and/or

- Predicted level of success, etc.

[0147] Fig. 18 illustrates how each module within a given range of possible‘journeys’ or ‘tracks’ may be assigned a relative percentage efficacy number or may be flagged as ineffectual or contra- indicated/causative of adverse effects because of morbidities or other factors, such that the system will aim to optimise based on path of fewest nodes and with highest payoff (where payoff is determined as risk-minimisation or predicted remediating effect).

[0148] This component may act as an engagement, adherence, and risk-reduction and remediation layer, surfacing goals and targeted interventions for users which serve to: - Reduce likelihood of preventable risk factors becoming morbidities;

- Maintain key health/life metrics;

- Stall further morbidities;

- Reverse specific conditions (e.g. T2D); and/or

- Provide targeted remediation interventions in the case of accidents or injuries (e.g. to expedite return to work).

[0149] Fig. 19 illustrates, by way of example, how as well as interventions being informed and suggested based on a user’s data points, risk stratification, and 2PLE, there may be a direct feedback loop into those components from the presentation layer by virtue of ongoing risk assessment, i.e. the achievement of system-suggested goals (such as“lose 1 2kg of weight and increase your Vo2 Max to 44.8 and lower your RHR to 62bpm”) is directly tied to specific risk-ratings, such that achieving the above may lead to either a maintenance of risk/2PLE score, reduction by 5% (and concomitant reduction of premium through continuous underwriting of $22 per month) or some such other configuration.

Example - Continuous underwriting

[0150] The present invention may, in an exemplary embodiment, leverage in-built risk models and 2PLE models to engage in a process of continuous member underwriting. Each user of the platform, by virtue of passive and active data collection during the use of the platform, may have their data points either progressively (i.e. incrementally as each data point is added by the user or ingested by integrated devices), or, be batch- processed by a personalised underwriting engine, which may be embedded within the health and wellbeing platform, or may, use a third party proprietary or‘commercial off the shelf (COTS) underwriting rules engine, with rules determined by the risk appetite and business model of the (re)-insurer or based on pre-defined tolerances in the risk models themselves.

[0151 ] This continuous underwriting may be used in order to determine and arrive at a variety of outcomes. By way of example, as shown in Fig.20, a continuous underwriting system may typically arrive at outcomes which may include a decision to provide insurance (and an associated price for said insurance-provision, if applicable), a decision not to provide insurance, or, that there is presently insufficient information to make a decision. [0152] Each of this scenarios is detailed, by way of example, as follows: a. Insufficient data

[0153] If there is insufficient data 201 to provide the user 25 with a life / health insurance classification (e.g.‘standard’,‘preferred’,‘super-preferred’, etc.), quotation or price, then the system may prioritise one or more outstanding data points based on the current completeness of the requisite data and display these to the user in order to elicit additional data.

[0154] This prioritisation of data capture may be in totality (whereby a user who has registered, for example, 8 of a requisite 10 data points to obtain a classification, quotation and / or price may be presented with the remaining 2 questions to complete), or, may be incremental (i.e. the remaining questions may be served up asynchronously and / or separately and / or at a later time).

[0155] Such data collection may be passive (i.e. automatically ingested from a third party connected device or system), or, active (whereby a user actively completes a given question, for example, or provides a data point to the system). b. Insurance offer

[0156] A classification, quotation or price 202 based on the user’s data points may alternatively be presented to the user 25 in the form of an insurance offer.

[0157] This may furthermore be supplemented by additional (conditional and / or goal- oriented) offers, for example, a pricing quotation based on the user’s existing health risk profile and data points and a supplementary offer based on the system’s determination of possible future target health risk profile or system-determined user goal (for example, feedback may be provided, such as“your term life insurance offer is based on your current profile and is $450 per annum; however, if you can reduce your RHR by 2.3 bpm, and increase your sleep to >7 hours per night within 2 months, your price will be $405 per annum”). c. No Insurance offer

[0158] A determination not to offer any insurance product 203 may alternatively be provided to the user. [0159] This may be because the (re)-insurer is not prepared / able to offer protection to the user 25 based on their current health / risk profile, avocation(s), insurance and / or credit history, financial standing, geographical location, the cost of supplementary (referred and / or medical), or because underwriting is too expensive or the pricing offer is not competitive in the market, etc.

[0160] As shown in Fig 21 , each user 25 of the platform undergoes a process of continuous underwriting, whereby the user is assigned a position on a spectrum from (nominally) 0 to 100, where, 0 either represents no data whatsoever or the highest possible health or other insurance risk (e.g. creditworthiness) or lack of eligibility (e.g. as a result of the user’s country of residence) as determined by the system based on the data and models contained therein (e.g. BMI of 40, self-reported history of 4+ heart attacks in the past 2 years, heavy smoker), and, 100 represents the‘best’ health profile and risk outlook and life expectancy, eligibility, and creditworthiness possible within the system.

[0161 ] Each user’s score and placing on the spectrum is subject to continuous fluctuation as their health profile and risk outlook changes and new data points are collected, such that a user with a score of 45 on day 10 may have a score of 38 by day 50.

[0162] Based on the user’s score, the following scenarios may typically be possible:

[0163] A user may not be offered any insurance product (area 1 on the spectrum shown in Fig 21 ). A user may not be offered any insurance product, but, may be placed into an intervention or treatment stream within the system which will target those specific risk factors which currently preclude them from being offered an insurance product, with a view to increasing their underlying score (area 2 on the spectrum Shown in Fig 21 ).

[0164] If the score is low because of a lack of requisite data points, the system may prioritise a specific sequence of (reflexive) data capture in an effort to understand their risk profile better and determine if their final, complete score is higher or lower.

[0165] A user may receive one or more insurance offers conditional on additional underwriting questions or disclosures, e.g. the user may be informed they may be eligible for one or more insurance products, but require a blood draw or telephone underwriting conversation or the provision of further data or medical evidence in order to confirm their score (area 3 on the spectrum shown in Fig 21 ).

[0166] A user may be offered an insurance offer and price, including one or more insurance products (e.g. term life, critical illness, travel insurance, etc.) with no additional underwriting questions required (‘straight-through processing’), with the option to proceed straight to purchase of the product (depicted as area 4 on the spectrum shown in Fig 21 ).

[0167] The system of the present may include an embedded, smart product

recommendation system which is underpinned by algorithms which use actively inputted or passively ingested data points and the risk classification engine and underwriting engine, as well as demographic, geographic, and financial data, amongst others, to recommend to a user one or more protection and / or health and wellbeing products, in addition to targeted content (‘Insights’) and signposting to healthcare

services/screenings, etc. based on the health risk profile, demographic, geographic, and other factors of the user 25.

[0168] For example, if the system determines from geo-location data that a user 25 is at an airport, and, has a health risk profile which can be underwritten and priced, and, meets standard eligibility criteria (country of origin, destination country/countries, age, etc.), the user may be offered a travel insurance product.

[0169] Alternatively or additionally, for example, if the system determines from social media data that the user has recently had a child and meets similar underwriting criteria for eligibility as set forth above, the user may be offered a term life product.

[0170] Alternatively or additionally, for example, if a user self-reports a family history of, or the system determines that the user is at an elevated risk of developing cancer, the user may be presented with targeted risk-reduction interventions, signposting (e.g. for breast cancer screening), and content to help reduce the user’s risk of developing cancer and may be offered a cancer protection product (e.g. to cover risks of lung, prostate, colorectal and breast cancer).

[0171 ] This approach can likewise be extended into other protection products, such as critical illness and medical expense cover, etc., as well as other health risks, such as diabetes, such as illustrated in Fig 22(a).

[0172] In Fig 22(a) is illustrated exemplary display screens which may be presented to a user to initiate user feedback and thereby ascertain a user’s risk of developing diabetes. This assessment may be done using of direct user information and/or in conjunction with, in this example, an in-build diabetes kit.

[0173] In Fig 22(b) is illustrated other exemplary display screens which may be presented to the user to initiate user feedback and thereby ascertain a user’s risk of as a risk of a heart or blood condition.

[0174] Thus, the overarching user’s journey within the system may broadly reflect the steps shown in Fig 23, including ongoing data collection, risk-profiling and continuous underwriting for pre-selection, provision of personalised health risk-reduction

interventions and content (e.g.“Because of your family history, age, and location, your risk of developing breast cancer over the course of your lifetime is x%, but if you can increase you intense exercise minutes to y mins per week and reduce your BMI by z points, you will reduce your risk by 1/3”), signposting to regular screenings, visits to specialists and / or GPs, and targeted and personally priced offers for specific insurance and health and wellbeing products based on the user’s profile (e.g. offers to purchase a blood pressure monitoring kit and / or offers to purchase life insurance). This illustrative user journey and its touchpoints with potential external data sources is further described in Fig 24.

[0175] Furthermore, to support increased insurance and health understanding, the system may provide behavioural nudges and boosts as part of the risk-reduction feedback which serve to raise awareness of users’ risks based on the system’s underlying risk models and user data. These may be rules-based (such as offer just-in- time content response to a user’s data input) or leverage Markov Chain or ANN and other deep learning / Al techniques to dynamically serve up content based on a range of correlations and patterns of user interaction, location, passive data, active data, and system algorithms.

[0176] Since it is well understood that insurance products are often complex financial instruments and may not be readily comprehensible to a system user, the system may also use active and passive data (e.g. financial transaction data, location, demographic information, proprietary profiling algorithms, etc.) in order to increase the user’s insurance and / or financial literacy based on their personal situation (e.g. age, marital status, financial status and spending habits, level of debt, etc.) and health risk factors as described above, as well as in-context nudges and‘boosts’ (which are“context-specific, individualized and efficient interventions into consumers' cognitive processes that aim at improving their decision-making competencies”).

[0177] These boosts may serve to provide real-time, contextual education and insight at the point of awareness and consideration of purchasing an insurance product in order to furnish the user with additional competences and decision-making capability, such that they more readily understand what they may be purchasing and why it may or may not be suitable for them.

[0178] These nudges and boosts may, by way of example, take the form as shown and described in Figs 25 and 26.

[0179] The present invention, as hereinbefore described, has a number of advantages over known products/methods.

[0180] The invention not only classifies health/fitness as single-cause modality; it determines risk holistically based on broad range of data points, ongoing data, and a feedback loop between the intervention and risk-classification, not providing generic goals (e.g. aim to walk 10,000 steps), but rather highly personalised goal to risk- reduction formulas, both individually and in combination with other goals and factors (e.g. an extra 1 ,232 steps over a month corresponds to CVD, stroke, T2D, and all-cause mortality risk-reduction of x).

[0181 ] Personalised, precision life expectancy is not only therefore more accurate at a point in time, but also, it is dynamically updated based on a user’s life and health stage and specific activities and incidents (e.g. sickness).

[0182] Taken together, this provides a far more accurate solution for underwriting purposes than, for example, generic health scores, such as those based on limited and static point-in-time data points, such as age, height, weight/BMI, smoking status, and gender which may be collected and priced at the commencement of the policy, but not updated again until the policy lapses. [0183] Improved (less costly, more accurate) underwriting vs. traditional health and life underwriting is therefore achieved because of continuous underwriting.

[0184] The present invention has many commercial uses/applications, including, but not limited to:

- licensing to insurers, brokers, and reinsurers to streamline underwriting;

- improve accuracy and transparency of pricing (health, life, travel) to

improve CX/reduce friction to increase GPW and lower cost of acquisition;

- increasing member acquisition and retention for

insurers/reinsurers/brokers;

- improving risk profile of at-risk and above (e.g. chronic conditions) policyholders, increasing length of time premiums will be paid and reducing costs of remediation by proactively managing risks in advance (preventive);

- the claims experience via e.g. auto-payouts of claims in event of hospitalisation or death of partner (based on data ingestion);

- improving cost profile of insureds by preventing risk and remediating risk via platform;

- improving the engagement and retention of prospective, new, and

existing policyholders;

- improving up-sell and cross-sell opportunities to policyholders based on life events;

- ability to create more custom insurance protection offers for consumers based on their needs (known and identified or not);

- improving understanding of risk pool of policyholders;

- opening up new markets and risk pools for insurers by broadening the definition of what is insurable risk and pricing based on anticipated risk- reduction levels achievable within the system;

- providing‘pay as you live’ insurance, which will:

- reduce premiums for individuals whose lifestyle is demonstrably reduced in risk;

- fluctuate based on improvements or deteriorations in health and fitness and risk;

- is modular, triggered based on life events (e.g. marriage, children, moving house, etc.); and, - is reflective and adaptive of all life stages and situations (pre and post claim, e.g. stages of RTW claim).

- providing the ability for insurers to adjust price and/or policy cover in real time (e.g. upon the system knowing a user has children and a mortgage to produce a targeted income protection offer); and,

- providing discounts/price maintenance, regardless of age, based on

achieving risk- reduction goals and key biometrics and psychosocial data points (beyond re-rating based on health questionnaires).

[0185] This system of the present invention therefore seeks to accurately and

comprehensively assess and stratify individuals’ risk on a personalised level, as well as providing innovative risk- reduction health journeys and interventions for people to drive down their risk of mortality, morbidity, YLL, and reduced quality of life. Ultimately this will allow global (re-)insurers to more accurately and fairly price risk, streamline

underwriting, ensure fairer and more open access to insurance, reduce financial stress and under-protection of individuals, and to lower the local and global burden of disease and preventable ill-health.

[0186] The present invention therefore seeks to provide a digital implementation and productisation of a series of risk stratification models and accompanying algorithms which leverage many more data points about an individual to inform their health and life risk, harnessing physiological, psychological, environmental, social, and other data in real time to accurately assess risk and continually underwrite, in contrast to standard underwriting models which look only at crude‘lag’ indicators of risk and largely self- reported biomarkers such as age, gender, and smoking status and underwrite statically (i.e. once only). This goes far beyond generic risk models and even cohort-level modelling: instead, it looks to classify all individuals to an‘n of T level of

personalisation, something which is truly unique in the market.

[0187] Furthermore, the risk-reduction and behaviour change platform herein described leverages this set of risk stratification algorithms to classify individuals and dynamically assign them personalised health journeys in order to drive down their risk levels (or maintain, if appropriate), as well as predicted outcomes (i.e. likelihood of reducing one or more risks by any given percentage amount and the associated remedial risk). [0188] Through a unique combination of evidence-based science, gamification, and behavioural psychology theory, this helps users reduce, arrest or reverse their risk of diabetes, CVD, mental health conditions and a range of other chronic diseases, help them manage their health better, manage existing claims scenarios (e.g. expedite recovery and rehabilitation in the event of absence from work as part of TPD/IP claim), and coordinate a personalised health journey spanning in-platform features, smart digital signposting of clinical/coaching/allied health services, and other risk-reduction services, making end to end healthcare frictionless for the consumer. This will not only support users’ own health accountability, but also drive down burden of disease costs, expedite return to work, and help people live healthier, happier, longer lives.

[0189] The present invention therefore provides a personalised health/wellbeing monitoring system and motivational system for users which seeks to improve the health of users. In particular, the present invention seeks to motivate users to improve their health for their own wellbeing. The present invention also has significant application to the insurance industry in being able to personalise insurance, and insurance needs, on an individual basis, rather than on the basis of an age group, or other broad grouping.

[0190] The present invention motivates the user to improve their health with the added benefit of reducing their premium on their medical, life or other insurance. Users of the present invention will therefore be rewarded with a reduction in their insurance premiums or the like and presented with cover options based on their own health needs, individual risks, and other data about the individual, such as known level of debt, financial transaction data, geographic location, and spending habits. This provides significant motivational incentive to users who achieve a reward by improvement in their health and by a financial incentive.

[0191 ] Wherever it is used, the word“comprising” is to be understood in its“open” sense, that is, in the sense of“including”, and thus not limited to its“closed” sense, that is the sense of“consisting only of”. A corresponding meaning is to be attributed to the corresponding words“comprise”,“comprised” and“comprises” where they appear.

[0192] It will be appreciated that numerous variations and modifications to the invention will become apparent to persons skilled in the arts. All such variations and modifications should be considered to fall within the spirit and scope of the invention is broadly hereinbefore described and is hereinafter claimed.