Privacy-realted context information for ubiquitous health

O iginal Pape P i acy-Rela ed Con ex In o ma ion o Ubiqui ous Heal h An o Seppälä, MS Comp Sc; Pi kko Nykänen, PhD; Pekka Ruo salainen, DSc (Tech) Cen e o In o ma ion and Sys ems, School o In o ma ion Sciences, Uni e si y o Tampe e, Tampe e, Finland Co esponding Au ho : An o Seppälä, MS Comp Sc Cen e o In o ma ion and Sys ems School o In o ma ion Sciences Uni e si y o Tampe e Kansle in inne 1 Tampe e, 33014 Finland Phone: 358 407069919 Fax: 358 32191001 Email: [email p o ec ed]i Abs ac Backg ound: Ubiqui ous heal h has been de ined as a dynamic ne wo k o in e connec ed sys ems. A sys em is composed o one o mo e in o ma ion sys ems, hei s akeholde s, and he en i onmen . These sys ems o e heal h se ices o indi iduals and hus implemen ubiqui ous compu ing. P i acy is he key challenge o ubiqui ous heal h because o au onomous p ocessing, ich con ex ual me ada a, lack o p ede ined us among pa icipan s, and he business objec i es. Addi ionally, egula ions and policies o s akeholde s may be unknown o he indi idual. Con ex -sensi i e p i acy policies a e needed o egula e in o ma ion p ocessing. Objec i e: Ou goal was o analyze p i acy- ela ed con ex in o ma ion and o de ine he co esponding componen s and hei p ope ies ha suppo p i acy managemen in ubiqui ous heal h. These p ope ies should desc ibe he p i acy issues o in o ma ion p ocessing. Wi h componen s and hei p ope ies, indi iduals can de ine con ex -awa e p i acy policies and se hei p i acy p e e ences ha can change in di e en in o ma ion-p ocessing si ua ions. Me hods: Scena ios and use s o ies a e used o analyze ypical ac i i ies in ubiqui ous heal h o iden i y main ac o s, goals, asks, and s akeholde s. Con ex a ises om an ac i i y and, he e o e, we can de e mine di e en si ua ions, se ices, and sys ems o iden i y p ope ies o p i acy- ela ed con ex in o ma ion in in o ma ion-p ocessing si ua ions. Resul s: P i acy- ela ed con ex in o ma ion componen s a e si ua ion, en i onmen , indi idual, in o ma ion echnology sys em, se ice, and s akeholde . Combining ou analyses and p e iously iden i ied cha ac e is ics o ubiqui ous heal h, mo e de ailed p ope ies o he componen s a e de ined. P ope ies de ine explici ly wha con ex in o ma ion o di e en componen s is needed o c ea e con ex -awa e p i acy policies ha can con ol, limi , and cons ain in o ma ion p ocessing. Wi h p ope ies, we can de ine, o example, how da a can be p ocessed o how componen s a e egula ed o in wha kind o en i onmen da a can be p ocessed. Conclusions: This s udy added o he ision o ubiqui ous heal h by analyzing in o ma ion p ocessing om he iewpoin o an indi idual’s p i acy. We lea ned ha heal h and wellness- ela ed ac i i ies may happen in se e al en i onmen s and si ua ions wi h mul iple s akeholde s, se ices, and sys ems. We ha e p o ided new knowledge ega ding p i acy- ela ed con ex in o ma ion and co esponding componen s by analyzing ypical ac i i ies in ubiqui ous heal h. Wi h he iden i ied componen s and hei p ope ies, indi iduals can de ine hei pe sonal p e e ences on in o ma ion p ocessing based on si ua ional in o ma ion, and p i acy se ices can cap u e p i acy- ela ed con ex o he in o ma ion-p ocessing si ua ion. (JMIR Mheal h Uheal h 2014;2(1):e12) doi:10.2196/mheal h.3123 KEYWORDS ubiqui ous heal h; p i acy; con ex in o ma ion; us ; policy JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.1h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X In oduc ion O e iew Ubiqui ous compu ing makes i possible o collec all kinds o da a anywhe e and any ime [1] and allows in eg a ion o heal h ca e deli e y and se ices in o people’s e e yday li es [2,3]. This pape builds on a concep ual amewo k [4] in which ubiqui ous heal h is de ined as an open and dynamic ubiqui ous in o ma ion space. The space is p esen ed as digi al sys ems ha consis o one o mo e in o ma ion sys ems, hei s akeholde s, and en i onmen s. These sys ems c ea e a dynamic ne wo k ha o e s and p o ides se ices o ci izens. In he in o ma ion space, indi iduals and se ice p o ide s can selec , ailo , and combine se ices and sys ems ha belong o he ne wo k. To enable access o pe sonal in o ma ion, indi iduals and p o ide s need o discuss us , p i acy le el, and p o e ed se ice. Ubiqui ous heal h se ices can be o e ed by p o ide s ha a e licensed and egula ed by medical e hical codes and heal h ca e–speci ic legisla ion and o he ju idical no ms and by ac o s ha a e no a ec ed by heal h ca e– ela ed egula ions. To sepa a e hese wo g oups, we di ided hem as egula ed heal h ca e se ices and o he se ices. P o ide s o e ing egula ed heal h ca e se ices ha e s ic de ined esponsibili ies and obliga ions conce ning se ice p o ision, ca e, p o essionals, documen a ion, and in o ma ion p ocessing. The e a e also gene al egula ions on p i acy and secu i y equi emen s (eg, da a p o ec ion and p ocessing di ec i es) and business domain–speci ic egula ions. Regula ions co e laws; no ms; good p ac ice guidelines; and o he ules con olling, cons aining, o limi ing ac i i y o pa icipan s. These egula ions can a ec ubiqui ous heal h se ices bu hey o en do no mee he challenges o echnological inno a ions well. In ubiqui ous heal h, us wo hiness and p i acy a e key challenges [4-6]. The e a e p i acy h ea s c ea ed by au onomous and hidden p ocessing o in o ma ion and ich con ex ual me ada a. The e is no p ede ined us be ween pa icipan s, and he business objec i es, needs, in e es s, and policies o s akeholde s may be unknown o he indi idual [4]. In o ma ion in ubiqui ous heal h is highly sensi i e and con iden ial, and he exis ence o se ices and ac o s ha a e no s ic ly egula ed by heal h ca e-speci ic legisla ion c ea es h ea s and isks o indi idual p i acy. In addi ion, in o ma ion p ocessing can happen in mul iple sys ems and si ua ions wi h di e en egula ions, and isks o seconda y use exis . The lack o p ede ined us and p i acy isks emphasizes he impo ance o an indi idual's abili y o con ol his o he p i acy. Fo us ed in o ma ion p ocessing in ubiqui ous heal h, we ollow he p inciples p esen ed in Ruo salainen e al [4] and acco ding o hem, an indi idual should ha e he igh o e i y dynamically he us wo hiness o he ubiqui ous heal h ne wo k and any sys em ha equi es o p ocesses he indi idual’s pe sonal in o ma ion o seconda y pu poses; con ol pe sonal heal h in o ma ion p ocessing, inside sys ems and be ween hem; be no i ied o all si ua ions and con ex s in which pe sonal in o ma ion is collec ed, p ocessed, s o ed, and/o disclosed; and c ea e si ua ion-speci ic, con ex -awa e, and g anula pe sonal p i acy and us policies, which con ol how pe sonal in o ma ion is collec ed, p ocessed, disclosed, sha ed, s o ed, o des oyed. Sys ems and s akeholde s should ha e he esponsibili y o ensu e us e i ica ion by publishing hei p i acy policies and en i onmen al and con ex ual ea u es; openness o in e es s, business needs, and policies as well as hei ela ionships wi h o he sys ems; and anspa ency o in o ma ion p ocessing. To p o ec his o he igh s, an indi idual needs in o ma ion abou p i acy, ha is, p i acy a ibu es, o de ine his o he pe sonal p i acy p e e ences. P i acy a ibu es enable p i acy o be a conc e e issue o indi iduals. In Nykänen e al [7], we de ined p i acy a ibu es as bene i , bene olence, capabili y, compe ence, con idence, con ex , eliabili y, and alue. P i acy a ibu es and hei con en s ha e no gene ally been esea ched widely. The ocus in his s udy is he con ex a ibu e, which e e s o he si ua ion in which da a a e c ea ed o p ocessed. The objec i e is o analyze and de ine p i acy- ela ed con ex in o ma ion componen s and hei co esponding p ope ies. When da a a e c ea ed, a con inuum o da a is bo n. Du ing he di e en p ocessing si ua ions, da a o i s p ope ies may change. O iginal con ex e e s o a si ua ion when da a a e c ea ed. In a ious use con ex s and p ocessing si ua ions, con ex in o ma ion is inc emen ally c ea ed and i desc ibes he cu en con ex and enables acking o he con ex his o y. Thus, da a ha e embedded con ex in o ma ion ha can be used by p i acy se ices o us calcula ion and o decide whe he p ocessing is allowed (Figu e 1). An indi idual’s p i acy p e e ences can be implemen ed wi h adap able p i acy policies. In p e ious wo k [8], we concluded a o mula o p i acy policies o con ain (1) us in o ma ion ha is a alue o a sys em- o en i onmen -speci ic calcula ion o egula o y compliance and us wo hiness; (2) sensi i i y o he da a; (3) si ua ion o he in o ma ion use; and (4) pu pose o he da a collec ion o use. Policy o mula ion is a decision p ocess in which an indi idual selec s p i acy ules and se ices and how much in o ma ion can be aded compa ed o he o e ed se ice and he le el o p i acy a ibu es. In his s udy, ou hypo hesis is ha con ex in o ma ion enables o mula ion o con ex -awa e p i acy policies hence enabling us wo hy p ocessing o pe sonal heal h and wellness in o ma ion and ealizing indi iduals’ igh s o p i acy in ubiqui ous heal h. In Ruo salainen e al [8], we p esen ed a p i acy a chi ec u e ha could use con ex in o ma ion in us calcula ion and in con ex -awa e p i acy policies o con ol an indi idual's pe sonal in o ma ion. Wi h his s udy, we add knowledge o ou ea lie esea ch by s udying he p i acy- ela ed con ex in o ma ion and by de ining he co esponding componen s and hei p ope ies ha suppo p i acy managemen in ubiqui ous heal h. JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.2h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X Figu e 1. Da a con inuum and con ex in o ma ion. P io Wo k P i acy and T us P i acy e e s o an indi idual’s abili y o con ol in o ma ion abou him- o he sel [9]. P i acy is a e y pe sonal concep and dependen on he con ex , because i may a y among ju isdic ions, cul u es, economies, ime, and indi iduals [10-12]. Smi h e al [13] claim ha p i acy is so bound o he speci ic con ex ha i canno be concep ualized as a single and unambiguous concep ; a he i should be ea ed as a se o in e es s. Cla ke [14] a gues ha i is use ul o unde s and p i acy as he in e es o keeping pe sonal space ee om in e ence and has di ided p i acy in o ou dimensions: pe son, pe sonal beha io , pe sonal communica ions, and pe sonal da a. In o ma ion p i acy means ha pe sonal in o ma ion should no gene ally be a ailable o o he pe sons o o ganiza ions and an indi idual should ha e majo con ol o in luence o e he pe sonal da a con olled by o he s and i s use [14]. In his esea ch, we e e o p i acy as an indi idual’s pe sonal iew wi hin he legisla i e bounda ies. T us is a concep closely ela ed o p i acy, and usually, he highe he alue o us , he lowe he need o p i acy [4]. T us implica es he willingness o sha e pe sonal in o ma ion wi h o he s [15]. Schoo man e al [16] emphasize ha us is based on a ela ionship and he le el o us exp esses he le el o isk an indi idual is willing o ake. Abdul-Rahman and Hailes [17] ha e de ined h ee cha ac e is ics o us : (1) us is subjec i e, (2) ac ions we canno moni o a ec us , and (3) us le el is dependen on how o he s’ac ions a ec ou ac ions. Se e al us models has been de eloped o calcula ing us wo hiness [16,18-20]. Ubiqui ous compu ing sys ems should be open and dynamic, because p e-iden i ica ion o pa icipan s is impossible and hey migh change egula ly [21]. In hese kinds o dis ibu ed en i onmen s, collabo a ion is i al because mul iple sys ems oge he y o achie e goals and pe o m asks and i is c ucial o sys ems o know which en i ies hey should o should no in e ac wi h [22]. T adi ional p i acy and secu i y solu ions a e no adequa e o ubiqui ous en i onmen s because he e is no cen al con ol o p ede ined use s o policies [19,21,23]. P i acy and secu i y a chi ec u e and decisions need o be based on us and i s p ope ies [19,21,24]. Con ex and Con ex Awa eness Con ex has been mos ly de ined wi h use p o ile, use emo ion, and use loca ion and iden i ies o nea by people and objec s and changes o hose objec s [25-28]. Acco ding o Dey e al [29], he h ee mos ele an en i ies a e places, people, and hings. These en i ies ha e o be conside ed om di e en iewpoin s such as loca ion, ac i i y, and iden i y. Dou ish [30] p oposes ha con ex and con en canno be sepa a ed; he con ex a ises om he ac i i y i sel and i canno be an ex e nal desc ip ion o he se ing. He claims ha con ex is a ela ional, in e ac ional p ope y be ween objec s and ac i i ies and he scope o ea u es mus be de ined dynamically [30]. Dey and Abowd ([28], pp. 3-4) de ined con ex as: “Con ex is any in o ma ion ha can be used o cha ac e ize he si ua ion o an en i y. An en i y is an indi idual, place, o objec ha is conside ed ele an o he in e ac ion be ween a use and an applica ion, including he use and applica ions hemsel es.” This de ini ion is open and i conside s ha any in o ma ion ha is ele an o in o ma ion p ocessing in a si ua ion can be used as a con ex . Con ex in o ma ion can, o example, be in o ma ion abou he use , de ice, en i onmen , o si ua ion. Thus, i is meaning ul o alk abou con ex ela ed o some hing ha exis s. The e a e h ee main uses o con ex in o ma ion [29]: (1) p esen ing in o ma ion and se ices o a use o using con ex o p opose ac ions o be pe o med, (2) execu ion o a se ice au oma ically on behal o he use , and (3) applica ions can ag con ex o in o ma ion o la e e ie al. In con ex -awa e compu ing, applica ions and sys ems a e able o pe cei e hei su oundings and en i onmen , adap acco ding o he con ex , and pe o m au onomously. Con ex awa eness e e s o adap abili y, which means ha applica ions and sys ems exploi pe cei ed con ex in o ma ion and adap hei beha io acco dingly [31]. In his iew, con ex in o ma ion is in o ma ion ha enables beha io modi ica ion based on his in o ma ion and i s ela ions. The sys ems, applica ions, and en i ies ha e o de ine he scope hemsel es. JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.3h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X Acco ding o Viswana han e al [32], he key poin o success ul ubiqui ous heal h is con ex awa eness, and he e a e al eady se e al con ex -awa e applica ions in he heal h and wellness domain. A lo o esea ch has been done o suppo pe sonalized ac ions and se ices in home ca e, ch onic disease managemen , and ambien assis ed li ing [33-37] wi h di e en pe sonalized heal h s a us, body senso ne wo ks, ac i i y o beha io moni o ing, decision suppo , and eminde applica ions [32,33,35-40]. In he hospi al en i onmen , many p o essionals a e e y agile, and con ex -awa e echnologies may help by pe sonalizing se ices o hem by loca ion, ime, and social con ex [41]. Acco ding o p e ious s udies [33,42], he e a e se e al expe imen s on con ex -awa e compu ing ha ha e been c ea ed in hospi al en i onmen s o imp o e pa ien eco d managemen , communica ion among p o essionals, and in o ma ion sha ing by including con ex awa eness in pa ien oom equipmen . Policies In ubiqui ous en i onmen s, p i acy equi emen s can be exp essed wi h policies. P i acy policy can be unde s ood as a pe sonal s a emen on p i acy. Wi h policies, indi iduals can se compu a ional ules explici ly s a ing hei pe sonal p i acy p e e ences on how hei in o ma ion can be p ocessed, used, disclosed, and sha ed [21,43,44]. Policies a e ypically exp essed wi h a policy language [45]. To enable pe sonal p i acy policies wi h compu a ional ules equi es de ini ion o p i acy a ibu es. P i acy policies can be implemen ed wi h se ing alues on p i acy a ibu es. Con ex -awa e policies based on con ex in o ma ion enable dynamic adap a ion o p i acy con ol s a egies and ailo ed p i acy decision suppo se ices. A echnique called s icky policy enables a aching policies in o da a o ensu e ha da a a e p ocessed acco ding o an indi idual’s wishes [44]. Beh ooz and De lic [46] p opose a con ex -awa e p i acy policy language based on wo design conside a ions: (1) si ua ions and p i acy ules a e de ined sepa a ely, and (2) a con ex eques o can be speci ied based on i s iden i y o social ela ionship o a use . These p inciples mean ha p i acy policies a e se o di e en si ua ions. Ghosh e al [47] p esen ed a seman ically ich policy-based sys em ha can eason on use ’s con ex and hus p o ec s a use ’s p i acy dynamically du ing un ime. Schaub e al [23] p esen ed a p i acy con ex model wi h h ee majo en i ies—use , use ’s en i onmen , and use ’s ac i i ies. Thei model akes in o accoun in o ma ion, physical, and e i o ial aspec s o p i acy. Bloun e al [48] p oposed a con ex -dependen policy model in which ield con ex con ains in o ma ion when condi ions o he policy a e alid. These alues may be om ei he he subjec o he eques o . Me hods Scena ios and Use S o ies Scena ios a e means o desc ibe he sys em’s in ended usage. Scena io-based design echniques p oduce desc ip ions o how people do hings and how hey can accomplish di e en asks wi h he sys em. Wi h scena ios, designe s can ind new ways o doing hings and new hings o do. Scena ios cap u e goals, en i ies, beha io al in o ma ion (eg, ac ions, ac i i ies, and e en s) and wha people a e ying o accomplish wi h he sys em [49,50]. They can also desc ibe di e en ela ed ac o s wi h hei own objec i es. Typically, scena ios ha e a plo ha consis s o se e al e en s, hings ha happen du ing ac i i ies, changes in he se ing, e c. Scena ios a e wo k-o ien ed analysis me hods; hus, hey a e sui able o ou pu poses, because we a e analyzing ypical ac i i ies o an indi idual in an ubiqui ous heal h en i onmen o ecognize he needs o con ex in o ma ion. In ou p e ious a icles, we analyzed p i acy h ea s and he p inciples o us ed in o ma ion p ocessing [4], de ined p i acy a ibu es [7], and analyzed he equi emen s o in o ma ion ha should be used in p i acy policy o mula ion and common h ea s and challenges conce ning p i acy in ubiqui ous heal h [8]. Ou p e ious esul s c ea ed he amewo k o he scena io de elopmen and analyses and o he equi emen s o con ex in o ma ion. In his esea ch, we c ea ed scena ios ha we e based on ma e ials collec ed in ou ea lie empi ical esea ch on pe sonal wellness [51-53] pe o med wi h ocus g oups and li e a u e s udies ocusing on heal h and wellness ac i i ies and echnical applica ions on ch onic disease managemen , sel -heal h managemen , ubiqui ous heal h, and wellness app oaches. Scena ios we e designed o cap u e he cha ac e is ics o di e en si ua ions, such as a gene al wellness managemen si ua ion wi hou any speci ic needs and a speci ic se ing wi h a ch onic disease. Wi h scena ios, we could iden i y a wide selec ion o ypical ac i i ies in ubiqui ous heal h. We i s c ea ed wo ex ual scena ios desc ibing he main ac o s, hei backg ounds, and cu en heal h and wellness si ua ions and nex , we de e mined he main goals, ac i i ies, and en i ies. Then, we u he di ided bo h scena ios in o 10 use s o ies ha desc ibed in mo e de ail he ac i i ies and se ices he indi iduals needed in hei si ua ions. Each use s o y ocuses on 1 ac i i y o a scena io and i is a sho ex ual and in o mal desc ip ion o a use case. Because con ex a ises om ac i i y [30] wi h he use s o ies, we could cap u e ac i i ies in ubiqui ous heal h o iden i y in o ma ion-p ocessing si ua ions and p i acy- ela ed con ex in o ma ion. A i s , scena ios desc ibed ypical wellness app oaches emphasizing se ices ha a e no egula ed by heal h ca e egula ions, o example, li es yle managemen and heal h- ela ed beha io s. The objec i e was o ecognize ac i i ies and en i ies ou side egula ed heal h ca e se ices. Then we app oached ch onic disease managemen scena ios wi h a ocus on iden i ying collabo a ion be ween egula ed heal h ca e se ices and pe sonal a emp s o manage heal h ou side he p o ide ne wo ks wi h o he se ices. These scena ios we e analyzed o ecognize ac i i ies and in o ma ion-p ocessing si ua ions. To summa ize, hese scena ios helped us o analyze he aspec s o wo di e en si ua ions in ubiqui ous heal h: (1) ubiqui ous heal h wi hou egula ed heal h ca e p o ide s’pa icipa ion; and (2) ubiqui ous heal h wi h egula ed heal h ca e, o example, se ice po olio is a combina ion o se ices p oduced by a egula ed heal h ca e p o ide (s) and o he heal h and wellness p o ide s. JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.4h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X An Example Scena io As an example, we p esen he ollowing scena io. Pe e is a 23-yea -old heal hy s uden who begins o eel i ed and ill and he decides o seek help om s uden heal h se ices. A e a ew es s and doc o isi s, Pe e is diagnosed wi h ype 2 diabe es melli us. F om now on, Pe e has o pay a en ion o his habi s and choices conce ning heal hy li ing o he i s ime in his li e. We di ided his scena io in o mo e de ailed use s o ies desc ibing ac i i ies ela ed o ch onic diseases in a ubiqui ous heal h en i onmen . In Table 1, we p esen an example analysis o a use s o y. In Table 2, we p esen a de ailed example o a single ac i i y in ubiqui ous heal h wi h i s ela ed p i acy conce ns, Pe e ’s policies, and he con ex in o ma ion ha a policy example equi es. Table 1. An example analysis o a use s o y in ch onic disease scena io. Use s o y 2.1: Pe e ecei es a medical de ice wi h senso s o manage and ca e o his disease and au oma ically measu e and moni o his condi ion. De ices can also au oma ically in o m his doc o abou he esul s and majo changes. Indi idual and in o ma ion con olle wi h igh s o p i acy, o con ol p ocessing and seconda y use o in o ma ion. Pe e can decide who can access da a c ea ed by he de ice. Pe e needs p i acy policies o con ol his own pe sonal heal h sys em (PHS) use and he in o ma ion i con ains. Role Da a is c ea ed in he senso s and ans e ed o PHS. PHS analyzes he in o ma ion and compa es i o pas in o ma ion. PHS in o ms Pe e ’s doc o abou a majo change in a alue. Doc o accesses he in o ma ion and makes a medical decision. Ac i i ies Anywhe e. No heal h ca e–speci ic egula ions conce ning he en i onmen . In o ma ion sha ing is based on Pe e ’s known consen and p i acy policies. All in o ma ion c ea ed by he ce i ied de ice is us ed. The de ice is egula ed by speci ic legisla ion (eg, he Eu opean Union di ec i e on medical de ices). In case o a majo change in measu e- men in o ma ion, egula ed heal h ca e se ice will pa icipa e and hen he en i onmen will be s ic ly egula ed by heal h ca e–speci ic egula ions. En i onmen Medical de ice, Pe e ’s own PHS and possibly elec onic heal h eco d sys em. Senso and measu emen da a is s o ed in PHS and Pe e ’s heal h eco ds a e in egula ed elec onic heal h eco d sys em. Pe e has o al con ol o e his PHS. In o ma ion sys ems Pe e , medical de ice, PHS, and licensed medical p o essional (doc o ) wi h esponsibili ies conce ning ca e and pa ien s p i acy S akeholde s Ce i ied medical de ice measu ing blood suga le els PHS diabe ic in o ma ion analysis Regula ed heal h ca e se ice ac i a ed by Pe e ’s PHS in case o a majo change in Pe e ’s measu emen alues Se ices Measu emen and moni o ing da a om senso s and medical de ice Heal h and wellness in o ma ion in PHS is con olled by Pe e . The medical in o ma ion is egula ed in heal h ca e o ganiza ion’s elec onic heal h eco d sys em. In o ma ion con en In o ma ion is c ea ed by a ce i ied medical de ice con olled by Pe e . The en i onmen does no ha e any speci ic domain egula ions. In o ma ion is in Pe e 's con ol and he has ull igh s o i . Pe e 's pe sonal con ex -awa e p i- acy policies a e he main sou ce o limi a ions and cons ain s on in o ma ion p ocessing. O iginal con ex o he in o ma- ion Pe e ’s PHS is a us ed in o ma ion sys em in his con ol so i has ull p ocessing igh s and can ac i a e o he se ices i needed ollowing Pe e ’s p i acy policies. Pe e has de ined in his policies ha di e en measu emen and senso da a is e y sensi i e and se s limi a ion o wha pu pose in o ma ion can be used. In o he cases, PHS canno g an access o in o ma ion wi hou Pe e ’s au ho iza ion. O he han egula ed heal h ca e, se ices ha e o sha e hei p inciples o in o ma ion p ocessing, secu i y and p i acy policies, and o wha pu pose hey wan o p ocess he in o ma ion. Requi emen s o con ex p op- e ies JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.5h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X Table 2. An example analysis o an ac i i y: da a is c ea ed in he senso s and ans e ed o he PHS. Requi ed con ex in o ma ion o policy 1Pe e ’s policiesP i acy challenges and h ea s [8] Si ua ion: ac i i y, p ocessing ype, ac o , a ge , in o - ma ion sensi i i y, and pu pose o p ocessing 1. Pe e hinks ha his kind o da a is highly pe - sonal and can only be accessed au oma ically by a heal h ca e p o essional pa icipa ing in Pe e ’s ca e se ice. Lack o awa eness En i onmen : gene al p i acy and secu i y egula ions, loca ion, and socie y 2. To use he da a, anspa ency o p ocessing is needed; he e o e, he p o ide has o publish de- ailed p i acy and secu i y policies and allow hi d-pa y audi ing. I is di icul o know how da a is used in he u u e Se ice: ype, ole, p o ide , loca ion, and objec i e3. To p e en seconda y use, copying da a is no allowed. I copying is equi ed, Pe e has o be no i ied and his known consen is equi ed. Rela ionships be ween sys ems may be unknown Indi idual: ole, igh s o con ol in o ma ion, ela ion o he ac i i y, con iden iali y equi emen s 4. Heal h ca e p o essionals a e no allowed o disclose da a wi hou Pe e ’s known consen . Po en ial seconda y use o in o ma ion S akeholde : iden i y, ype, ole, pu pose, and jus i ica- ion o p ocessing Use s wan o con ol how sys ems use pe sonal heal h in o ma ion IT sys em: iden i y, ype, con olle How o gua an ee ha da a is p ocessed ollowing he legal cons ain s and acco d- ing o he indi idual’s policies Resul s In an open and dynamic ubiqui ous heal h in o ma ion space, he e a e no possibili ies o p ede ine en i ies o ac i i ies and mos aspec s o in o ma ion p ocessing a e dynamic. In he scena ios and use s o y analyses, we ecognized how di e en ac i i ies a e easons o in o ma ion-p ocessing si ua ions in ubiqui ous heal h, how se e al en i ies can c ea e and use in o ma ion, and how he same in o ma ion can be used la e o suppo di e en ac i i ies. In addi ion, scena ios showed ha ac i i ies could happen au onomously wi h in o ma ion sys ems e en wi hou human pa icipa ion; o ins ance, based on some measu emen o i al signs o moni o ing o da a. Thus, in o ma ion p ocessing happens because some en i y pe o ms an ac i i y in a ce ain en i onmen . Si ua ion desc ibes his occu ence and he e o e is chosen as he co e componen de ining p i acy- ela ed con ex in o ma ion. I is linked o a ce ain ac i i y; ha is, he eason o in o ma ion p ocessing. Con ex in o ma ion needs o include he whole si ua ion and all pa icipan s because o he dynamic na u e and limi a ions in p ede ining ac i i ies and s akeholde s in ubiqui ous heal h. As a esul o ou scena ios and use s o ies, we p esen he wo kinds o basic models o ubiqui ous heal h: ubiqui ous heal h wi hou egula ed heal h ca e p o ide s, and ubiqui ous heal h wi h egula ed heal h ca e se ice p o ide s. The i s case is an open en i onmen wi h mul iple en i ies wi h di e en kinds o domain en i onmen s and in e es s. All pa icipan s a e by de ini ion un us ed. Heal h ca e–speci ic egula ions do no apply, bu egula p i acy and secu i y legisla ions se limi a ions o in o ma ion p ocessing. In addi ion, di e en domains may ha e hei speci ic legisla ions (eg, social ca e, wellness se ices, medical de ices, o pha macy). En i onmen and en i y-speci ic egula ions and an indi idual’s pe sonal con ex wi h p i acy p e e ences a e necessa y o adap able p i acy policies. An indi idual’s ole, en i onmen , and p i acy equi emen s may a y be ween used se ices o in o ma ion sys ems and in o ma ion sensi i i y in luences hea ily on pe sonal policies. An indi idual’s igh s o con ol da a and in o ma ion mus be discussed wi h se ice p o ide s. In he second case, he e a e also en i ies ha a e a ec ed by heal h ca e–speci ic egula ions. Depending on who o wha p o ides se ice and/o con ols in o ma ion, he e migh be s ic heal h ca e–speci ic egula ions o se ice p o ision, o ganiza ions, p o essionals, in o ma ion sys ems, and in o ma ion p ocessing. Regula ed heal h ca e se ices a e o some ex en us ed and p i acy h ea s and isks occu especially when in o ma ion is ans e ed om hem o p ocessed beyond hei au ho i y. I is e y c i ical o cap u e who is esponsible o wha , whe e and how se ices a e p o ided, wha in o ma ion and sou ces a e used, how sensi i e he in o ma ion is, and who con ols pa icipa ing in o ma ion sys ems. In a p e ious s udy [4], we de ined ubiqui ous heal h o be composed o se ices, in o ma ion sys ems, s akeholde s, and hei en i onmen s. In addi ion o hese, we ha e o cap u e he con ex s o he in o ma ion-p ocessing si ua ion and i s objec and/o subjec . We should cap u e he ollowing componen s and hei p ope ies on p i acy, egula ions, and equi emen s o us ed in o ma ion p ocessing: wha happens (si ua ion); who is he subjec o he objec (indi idual); wha se ices a e ela ed o he si ua ion (se ice); whe e his si ua ion happens (en i onmen ); wha social ac o s a e ac i e in he si ua ion (s akeholde ); and wha compu a ional en i ies pa icipa e (IT sys em). In his esea ch, he p ope ies o he p i acy- ela ed con ex in o ma ion componen s and hei p ope ies a e de i ed by combining he esul s o he scena io analyses and he p inciples and equi emen s p esen ed in he ea lie esea ch. We analyzed he esul s o he scena io analyses o explica e conc e e p ope ies o ou componen s. In he example, we de i ed he con ex in o ma ion ha is needed o ul il he equi emen s o JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.6h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X policies and o minimize known p i acy h ea s. In his example, policy 1 in he Table 2 means ha Pe e se s a gene al policy ha da a c ea ed by senso s is highly sensi i e and can only be au oma ically accessed by heal h ca e p o essionals pa icipa ing in his ca e. Pe e has o al con ol o e his da a and he u u e use o da a is based solely on Pe e ’s wishes. The si ua ion occu s when a egula ed heal h ca e p o essional ies o access Pe e ’s da a o suppo Pe e ’s ca e and o ollow his condi ion. To manage his p i acy, Pe e needs in o ma ion abou he da a use ’s en i onmen and p ocessing wishes. I pa ies o he han a heal h ca e p o essional in Finland aking pa in Pe e ’s ca e se ice wan o access he da a, Pe e ’s known consen is equi ed. The da a use is a egula ed heal h ca e p o essional in Finland; ha is, p ede ined as somewha us ed and he/she can use he in o ma ion only o make medical decisions and o ollow Pe e ’s condi ion. The da a can only be accessed wi hin Pe e ’s PHS and he da a canno be copied o dis ibu ed. F om he example, we can see how Pe e needs se e al kinds o con ex in o ma ion o c ea e he example policy. F om he scena io analyses, we ha e de ined he p ope ies ha a e needed o ul il he p inciples o us ed in o ma ion p ocessing and equi emen s se o p i acy o mula ion conce ning con ex in o ma ion (Figu e 2). Figu e 2. P i acy- ela ed con ex in o ma ion componen s and hei p ope ies o ubiqui ous heal h. A si ua ion desc ibes in o ma ion p ocessing ha happens in a ce ain con ex because o some ac i i y and by/ o a ce ain indi idual. F om he scena ios, we lea ned ha en i onmen s migh a y a lo ; he e o e, we need o unde s and he en i onmen whe e he si ua ion happens and componen -speci ic en i onmen s (eg, indi idual, se ices, s akeholde s, and IT sys ems) o cap u e all p i acy aspec s. Wi h he en i onmen , we do no only mean loca ion and o he posi ion-based in o ma ion, bu especially impo an is o cap u e he ype o en i onmen . We ha e o pe cei e he p ope ies o en i onmen such as p i acy, secu i y, us - ela ed in o ma ion, and in o ma ion-p ocessing ules and esponsibili ies. Regula ions may di e a lo be ween en i onmen s and di e en businesses a e a ec ed by hei speci ic legisla ion. Cap u ing en i onmen is c ucial because echnological ad ancemen s such as cloud compu ing and big da a c ea e new ypes o p i acy isks. Fo example, i a se ice is o e ed in he Eu opean Union bu he da a a e s o ed o p ocessed in an in o ma ion sys em loca ed in he Uni ed S a es, he e a e di e ences in legisla ions conce ning p i acy, secu i y, o seconda y use o da a. People should be able o con ol whe e and why hei da a a e p ocessed. An indi idual componen desc ibes he ac ual subjec and/o objec o heal h and wellness ac i i ies in ubiqui ous heal h. I is linked di e en ly o si ua ions; an indi idual can c ea e hem, pa icipa e in hem, and/o is an objec . P ope ies needed om he indi idual a e he ole he/she has in he si ua ion, loca ion, and en i onmen and wha ela ion he/she has wi h he ac i i y. Also, an indi idual’s igh s o con olling in o ma ion p ocessing (eg, con en , disclosu e and access o in o ma ion), p i acy policies, sensi i i y and con iden iali y equi emen s and wha is his/he ela ion (eg, owne , con olle , o subjec ) o in o ma ion should be acknowledged. All hese hings a ec how and on wha basis sys ems can p ocess in o ma ion. A se ice componen desc ibes egula ed heal h ca e se ices and/o o he se ices ha can be o e ed by IT sys ems and/o s akeholde s. An IT sys em componen e e s o all compu a ional en i ies, which can include heal h in o ma ion sys ems, pe sonal heal h sys ems, ubiqui ous sys ems, de ices, senso s, e c. IT sys ems should be open abou hei p ocesses and publish hei p i acy and secu i y policies including how an indi idual’s p i acy is p o ec ed, ele ancy o p ocessing and ac ual da a p o ec ion speci ica ions, and de ailed in o ma ion-p ocessing p inciples. This would imp o e anspa ency o in o ma ion p ocessing and inc ease us wo hiness. I an IT sys em does no publish necessa y in o ma ion, his has o be cap u ed in he con ex in o ma ion. Because in o ma ion p ocessing can happen anywhe e, i is i al o cap u e i s con ex because he e a e se e al cha ac e is ics a ec ing p i acy ha may di e be ween IT sys ems; o example, ype, loca ion, o egula i e backg ound. Fo example, he e a e big di e ences in egula ions among in o ma ion JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.7h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X sys ems, egula ed medical de ices (eg, ha e o be ce i ied), and wellness de ices. S akeholde is he social componen desc ibing o ganiza ions and possible human pa icipan s. They can be ac o s o in e es ed pa ies in a si ua ion. They o e , pa icipa e, o a e in e es ed in se ices o e ed o indi iduals. Ou componen s can be used o inc ease us wo hiness o in o ma ion p ocessing because p i acy policies can be adap able and based on cons ain s, limi a ions, ules, igh s, and esponsibili ies se wi h si ua ional in o ma ion. Componen s can also be used o analyze ha he in o ma ion p ocessing ollows he p e e ences se by an indi idual’s p i acy policies and he equi emen s om he o iginal con ex o he in o ma ion. Fo example, in ou use s o y Pe e may disclose medical and li es yle da a o a se ice p o ide o ecei e a selec ed se ice. Pe e has se p i acy policies using p i acy p ope ies. Be o e disclosu e, Pe e (his p i acy a chi ec u e) needs he con ex in o ma ion om he se ice p o ide o calcula e i p ocessing is acco ding o he equi emen s se by Pe e ’s own con ex -awa e p i acy policies and he o iginal con ex o he in o ma ion. P i acy a chi ec u e can hen con i m ha he use con ex is alid acco ding o Pe e ’s pe sonal p e e ences and allow access o he in o ma ion (Figu e 3). Ou hypo hesis was ha p i acy- ela ed con ex in o ma ion could be used o o mula e con ex -awa e p i acy policies hence enabling us ed p ocessing o pe sonal heal h and wellness in o ma ion. In his s udy, we analyzed con en s o a p i acy a ibu e con ex and p esen ed componen s and hei p ope ies ha can be used as pa o p i acy policies by se ing si ua ional cons ain s and limi a ions. These cha ac e is ics a e also needed o cap u e in o ma ion-p ocessing con ex s om he p i acy pe spec i e. All componen s o p ope ies a e no necessa ily needed in all si ua ions. In addi ion, i some sys ems e use o coope a e in publishing con ex in o ma ion, his has o be cap u ed and acknowledged. Figu e 3. The use o p i acy- ela ed con ex in o ma ion in ubiqui ous heal h. Discussion In his esea ch, we p esen an app oach using p i acy- ela ed con ex in o ma ion o p i acy p o ec ion in ubiqui ous heal h. P i acy is a business-enable because indi iduals will no use hese se ices i hey canno manage hei p i acy and us . People need simple ools o manage hei p i acy and we ha e s a ed his by de ining he componen s si ua ion, en i onmen , indi idual, se ice, s akeholde , IT sys em, and hei p ope ies. These componen s desc ibe he c ucial p i acy- ela ed con ex in o ma ion needed o imp o e us wo hiness o ubiqui ous heal h. We p esen new knowledge by de ining con ex , which is one o he main p i acy a ibu es used in p i acy policy de ini ion. The esul s o his s udy can be used as a basis o c ea e mo e o mal models de ining p i acy- ela ed con ex in o ma ion in a compu e -unde s andable o ma . Ou esul s a e in line wi h he p e e ed p i acy le el model by Lede e e al [11] bu we ha e aken i a s ep u he and di ided con ex in o o iginal and use con ex and de ined mo e de ailed and conc e e p ope ies ha could be alued and measu ed and used by p i acy a chi ec u e o us calcula ion. Ubiqui ous heal h is s ill an eme ging ield combining highly egula ed heal h ca e wi h pe sonal heal h and wellness se ices and sys ems. In heal h ca e, legisla ion and egula ions de ine wha p i acy is and wha kind o igh s indi iduals ha e; ha is, p i acy is a s a e-de ined p ope y. Conside ing se ices and sys ems ou side he egula ed heal h ca e p i acy is a pe sonal p ope y o an indi idual; ha is, ee will. The indi idual has he igh o choose he use o his/he in o ma ion and de ine policies as o how, whe e, and o wha ex en he in o ma ion can be p ocessed. In ubiqui ous heal h, a p i acy model is a combina ion o hese wo models and can be con olled wi h JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.8h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X policies. Policies can be pe sonal p e e ences o de ined by egula ions. Using he scena ios, we could iden i y si ua ions ou side egula ed heal h ca e o ecognize equi emen s and cha ac e is ics o ubiqui ous heal h. Wi h o ganiza ion-cen ic heal h ca e p ocesses o wo k lows, we canno eally model ubiqui ous heal h as a whole because he e a e many se ices and sys ems wi hou p ede ined and egula ed p ocesses o wo k lows. In ubiqui ous heal h, se ice p o ision is based on cus ome ela ionships and ading on bene i s o se ices agains educing pe sonal p i acy. Indi iduals should be able o e i y he us wo hiness o se ice p o ide s and decide i hey a e p epa ed o disclose pe sonal in o ma ion and educe p i acy. Because se ices a e o en o e ed as dis ibu ed, pe sonalized and e en au onomous, he p i acy a chi ec u e should o e au oma ic p i acy se ices and adap dynamically o he si ua ion. Scena ios and use s o ies showed ha ubiqui ous heal h is mul idimensional wi h limi a ions o p ede ining si ua ions. The amoun o in o ma ion needed and c ea ed in hese si ua ions can be huge, and he con en and i s sensi i i y a y depending on he ac i i y pe o med. Ubiqui ous heal h is an open, dynamic, and collabo a i e en i onmen and p i acy needs o be based on us and i s p ope ies [19,21,24]. In heal h ca e, p i acy is mainly p o ec ed wi h access con ol and consen managemen . Access con ol is me ely one ool o p o ec p i acy. Managing p i acy in ubiqui ous heal h is a much b oade issue han jus con olling heal h ca e p o essionals’ access o da a. Access con ol wi h p ede ined igh s, oles, and consen s canno eally unc ion because he e is no cen al con ol o necessa ily p ede ined p ocesses, si ua ions, o ac o s. To ensu e p i acy in ubiqui ous compu ing, access con ol should be dynamic because o mul iple changing en i ies. Con ex in o ma ion enables dynamic managemen o igh s [54]. Consen is an example o a pe sonal policy bu in ubiqui ous heal h, policies a e needed o co e se e al di e en si ua ions ha a e mo e complex han hose ha consen s a e designed o . Policies ha e o be dynamic and con ex -awa e. Co adi e al [55] p esen a dynamic and lexible secu i y middlewa e ha uses con ex as a basic concep in secu i y policy speci ica ion and pe missions a e linked o he con ex s ins ead o use iden i ies o oles. Mos esea ch on p i acy o con ex -awa e compu ing ocuses on cap u ing use ’s con ex o ce ain ac o s and using ha in o ma ion o adap o p i acy p e e ences [23,47,54]. In his s udy, we ollowed he app oach o Beh ooz and De lic [46] o sepa a e si ua ions and p i acy ules. We iden i ied he necessa y in o ma ion o cap u e p i acy aspec s in in o ma ion-p ocessing si ua ions. Then, p i acy a chi ec u e can cap u e he si ua ion and he condi ions whe e da a a e c ea ed; ha is, he o iginal con ex and combine ha wi h indi idual's policies and con ol u u e use con ex s such as how, whe e, and by whom he in o ma ion can be used. Ou app oach needs in o ma ion om pa icipa ing sys ems and cu en ly i s a ailabili y depends on he goodwill o pa icipan s. Addi ional o his in o ma ion, p i acy a chi ec u e can use ex e nal sou ces o es ima ing us wo hiness o sys ems (eg, ecommenda ions om o he s, his o y, us alues, and us calcula ions). In he Eu opean Union, o ganiza ions a e equi ed o in o m indi iduals abou use o hei da a and publish p i acy policies ha should be comp ehensi e wi h high-le el desc ip ions o hei p i acy p ac ices [43]; howe e , hese a e no enough o sa egua d indi iduals’ igh s. These p i acy policies do no gene ally conside how da a a e ac ually p ocessed a e collec ion. So, one o he main challenges in p i acy p o ec ion is how o en o ce all ele an pa ies o explica e hei de ailed p i acy policies [43]. Cu en legisla ion is no ully p epa ed o handle p i acy h ea s o ubiqui ous compu ing and does no obliga e o ganiza ions o disclose hei de ailed p i acy policies o in o ma ion-p ocessing p inciples. In he u u e, legisla ion needs o include he needs o p i acy, ci izens’ igh s, and ubiqui ous compu ing. Ci izens ha e o be able o con ol p ocessing and seconda y use o hei pe sonal in o ma ion. Fu u e p i acy p inciples and no ms need o p og ess om high-le el p inciples o de ailed egula ions conce ning he p ocessing and use o in o ma ion. This would b ing openness and anspa ency o in o ma ion p ocessing and new kinds o esponsibili ies o o ganiza ions and in o med igh s o ci izens. In addi ion, au ho i ies o ce i ica e o ganiza ions should be able o audi p o ide s and o e ecommenda ions abou hei us wo hiness. The componen s de ined in his esea ch may ha e some limi a ions and may no be conclusi e; howe e , based on he scena io analyses hese a e needed. In addi ion, some p ope ies a e ha d o de ine explici ly o in measu able o ma . They ha e o be analyzed in mo e de ail and o mal models a e needed o implemen hem in compu a ional o ma . Also, we need mo e de ailed analysis o wha o ganiza ions should publish abou hei p ocesses and p i acy and secu i y policies and p inciples. To c ea e con ex -awa e p i acy se ices and policies in p ac ice, we need o de elop on ologies ha explica e componen s, p ope ies, and equi emen s ha we ha e p esen ed in his esea ch. On ologies a e o mal ep esen a ions and should co e di e en ac i i ies, se ices, IT sys ems, s akeholde s, in o ma ion con en , and especially ele an egula i e en i onmen s. Wi h on ologies, we can c ea e compu a ional ules ha can be used o en o ce egula ions and pe sonal policies in o ubiqui ous applica ions. Because i is p ac ically impossible o indi iduals o e alua e he us wo hiness o a sys em, and o unde s and de ailed p i acy and secu i y equi emen s and se pe sonal policies, we de eloped us -based p i acy managemen a chi ec u e o ubiqui ous heal h [8]. This a chi ec u e model desc ibes wha p i acy and secu i y se ices a e needed o enable us ed in o ma ion p ocessing in ubiqui ous heal h. The a chi ec u e will apply p i acy- ela ed con ex in o ma ion o c ea e p i acy and secu i y policies ha will ensu e ha in o ma ion p ocessing will no happen agains he wishes o he indi idual and he o iginal con ex o he da a. The a chi ec u e con ains decision suppo and policy se ices o indi iduals o help hem de ine pe sonal policies. This esea ch adds o he a chi ec u e model by de ining he equi ed p i acy- ela ed con ex in o ma ion componen s and hei p ope ies ha a e needed o c ea e implemen able ools and means o indi iduals o manage pe sonal in o ma ion p i acy. JMIR Mheal h Uheal h 2014 | ol. 2 | iss. 1 | e12 | p.9h p://mheal h.jmi .o g/2014/1/e12/ (page numbe no o ci a ion pu poses) Seppälä e alJMIR MHEALTH AND UHEALTH XSL • FO Rende X