Privacy in Mobile Health Applications for Breast Cancer Patients

P i acy in Mobile Heal h Applica ions o B eas Cance Pa ien s Jaime Benjumea, En ique Do onzo o, Jo ge Rope o, Oc a io Ri e a-Rome o, Alejand o Ca asco Uni e sidad de Se illa Depa men o Elec onic Technology Se ille, Spain [email p o ec ed], [email p o ec ed], [email p o ec ed], [email p o ec ed], [email p o ec ed] Abs ac —P i acy is a majo conce n o b eas cance pa ien s. When pa ien s use mobile heal h applica ions (mHeal h apps), many sensi i e da a a e handled by he applica ion de elope s. Gene al Da a P o ec ion Regula ion (GDPR) a ises as a solu ion o p i acy issues. In his pape , we analyze he p i acy policy o a sample o mHeal h apps o b eas cance pa ien s, de eloping a scale o check i GDPR is complied. Despi e p i acy is a key ac o in he adop ion o he use o mHeal h apps, he low le el o compliance wi h he GDPR o he analyzed applica ions was qui e su p ising. Thus, applica ion de elope s mus be conce ned abou his ma e . Keywo ds: b eas cance , p i acy, GDPR; mHeal h; mobile applica ions. I. INTRODUCTION B eas cance (BC) is he mos common cance in women [1,2]. Nowadays, he ea ly BC de ec ion and ad ancemen s in ea men s ha e esul ed in an inc eased su i o ship [3- 5]. Howe e , BC su i o s li e wi h he e ec s o hose ea men s and possible complica ions [6], equi ing hem o be p oac i e in he ca e o hei heal h. They mus add ess medical and psychological conce ns and needs in he pos - ea men pe iod [7]. Technological solu ions may p o ide e ec i e ools o heal h educa ion, sel -moni o ing, communica ion, e c., suppo ing hem in he sel - managemen o hose su i o ship issues. Connec ed Heal h, whe e hese solu ions may be included, is a echnology-enabled heal hca e deli e y model aimed o maximize heal hca e esou ces [8]. Mobile Heal h (mHeal h), a subdomain o Connec ed Heal h, is de ined as he deli e y o heal hca e o heal h ela ed se ices using po able de ices [9]. Among po able de ices, sma phones a e pa icula ly in e es ing due o hei inc eased global use, ubiqui y, high cos -e ec i eness, and capabili ies, such as acking use s’ beha io s and p o iding hem eal- ime eedback. The use o mobile heal h applica ions (mHeal h apps) o heal h and well-being p omo ion has inc eased apidly in ecen yea s [10]. In he case o BC, Giun i e al. e iewed BC mHeal h apps a ailable in he leading sma phones app s o es (Apple S o e and Google Play) inding 454 mHeal h apps in ended o pa ien s [11]. 114 o hose mHeal h apps we e ocused on disease managemen . Common ea u es included in hose BC mHeal h apps o disease managemen such as ac i i y acking, heal h dia ies, o pa ien epo ed ou comes, equi e pa ien o en e da a in o he app. Some o hose mHeal h apps e en allow pa ien s o sha e hei da a wi h o he s. Mos o BC pa ien s a e willing o sha e hei da a wi h a heal hca e eam, bu hey may be eluc an o sha e hem wi h o he s [12-14]. Da a p i acy is a common conce n among pa ien s ega ding he use o mHeal h apps o sel -manage hei heal h [15, 16]. P i acy and secu i y o use in o ma ion con ained in he mHeal h apps has been epo ed as a ele an issue in he e alua ion o hei sa e y and quali y [17]. In his pape , we assess he p i acy policies o a sample o BC mHeal h apps. Wi h his aim, we in oduce a no el scale o assessing p i acy. The scale is based on he Gene al Da a P o ec ion Regula ion (GDPR), applicable since 2018 ac oss he Eu opean Union (EU). Ou scale de ines a sco e o e e y mHeal h app based on se e al GDPR i ems ha mus be complied. We analyzed a sample o 9 mHeal h apps o BC sel -managemen , applying ou no el p i acy assessmen scale. The pape is o ganized as ollows. Sec ion 2 deals wi h he legal and echnical backg ound. Legal backg ound includes some GDPR ea u es and de ini ions ha a e use ul o ou esea ch. Resea ch backg ound p esen s a b ie desc ip ion o ela ed wo k p e iously published in bo h gene al and eHeal h p i acy policy assessmen s. Sec ion 3 desc ibes he p i acy scale designed based on i s p i acy policy and we de ine he sc eening me hod o selec ion o he BC mHeal h apps. In Sec ion 4, we show he main esul s o ou s udy. Finally, Sec ion 5 shows ou main conclusions abou he analysis. II. BACKGROUND A. Legal backg ound Regula ion 2016/679 o he Eu opean Pa liamen and o he Council, also known as Gene al Da a P o ec ion Regula ion (GDPR) [18], was published in he O icial Jou nal o he Eu opean Union on 27 Ap il 2016. GDPR is applicable o all membe s a es since 25 May 2018. GDPR ha monizes he legisla ion ac oss he whole Eu opean Union. I is an 88-page documen and con ains 99 a icles. Some o he mos impo an ea u es a e he ollowing: •Ha moniza ion: I applies o all Membe S a es di ec ly, wi hou needing o be ansla ed i in o a na ional law. •I applies o all he es ablishmen s in he EU, con olle s o p ocesso s – see de ini ions below -, e en i p ocessing occu s ou side he EU. (A icle 3.1). •I applies o EU people’s da a, e en i he con olle o p ocesso is no om he EU, bu i o e s goods o se ices o people in he EU (A icle 3.2) o i moni o s use s’ beha io s. •Accoun abili y: he con olle shall be esponsible o he compliance wi h GDPR. GDPR allows a con olle o adhe e o codes o conduc o ce i ica ion, so ha he compliance can be easily e i ied. (A icle 5.2) •Coe cion: Coe ci e measu es a e s onge . A supe iso y au ho i y may ine he con olle wi h up o 20,000,000 EUR. Also, a supe iso y au ho i y may adop p o isional measu es in case he e is an u gen need o ac o p o ec da a subjec s’ igh s. (A icles 83 and 66) Mo eo e , GDPR de ines some impo an concep s o egula ion. Mos de ini ions can be ound in a icle 4, excep o he wise indica ed. We a e using he ollowing in his pape : •Da a subjec : iden i ied o iden i iable na u al pe son whose pe sonal da a a e being p ocessed. Any da a ha makes someone iden i ied o iden i iable is pe sonal da a, acco ding o GDPR. •Con olle : “ he na u al o legal pe son, public au ho i y, agency o o he body which de e mines he pu poses and means o he p ocessing o pe sonal da a” •P ocesso : “a na u al o legal pe son, public au ho i y, agency o o he body which p ocesses pe sonal da a on behal o he con olle ” •Recipien : is o whom pe sonal da a a e eleased. This includes p ocesso s. •Thi d pa y: is someone di e en om he da a subjec , con olle o p ocesso who is au ho ized o p ocess pe sonal da a. •Rep esen a i e: a na u al o legal pe son designa ed by a con olle o p ocesso , no es ablished wi hin he EU. •Da a P o ec ion O ice (DPO): is a pe son ha mus be designa ed unde some ci cums ances (A icle 37). Wi hin i s du ies, DPO ad ise he con olle o he p ocesso , and moni o compliance wi h GDPR. (A icle 39). B. Resea ch backg ound The assessmen o p i acy policies has been a big conce n in ecen yea s in he heal h domain, pa icula ly in mHeal h due o i s apid g ow h. Con issa e al. [19] p esen a GDPR-based me hodology o assess p i acy policies. This me hodology checks i a p i acy policy includes all he in o ma ion equi ed by he a icles 13 and 14 o he GDPR, i i is ai , and i i can be easily unde s ood. The pape assesses 14 online pla o ms using his me hodology and ies o au oma e he assessmen . Howe e , hey do no de elop a scale. GDPR’s equi emen s and ecommenda ions om he In o ma ion Commissione ’s O ice o Uni ed Kingdom a e used by [20] o design a p i acy policy based on icons and highligh ed ex (called “GDPR label”). Thei pu pose is o compa e a ex -only GDPR wi h a GDPR label e sion. The labels cap u e all he in o ma ion ha is equi ed o be included in p i acy policies. This su ey sugges s ha GDPR label e sion is p e e ed o ex -only e sion. Machine lea ning and na u al language p ocessing echniques a e used in [21] o classi y p i acy policies. They use di e en GDPR a icles o build a lis o i ems o be assessed. A le el o isk is associa ed o each i em, depending on how a p i acy aspec is add essed by he p i acy policy. They ha e de eloped an applica ion ha ecei es a p i acy policy as an inpu and hen summa ize his policy using a g aphical use in e ace, so ha use s can easily iden i y po en ial isks in he p i acy policy. Renaud and Shephe d [22] make an exhaus i e s a e-o - he-a esea ch and a syn hesis o GDPR equi emen s o p o ide a guide o w i e p i acy policies. They ocus on usabili y. The a icles abo e a e e e ed o gene al p i acy. Howe e , p i acy is especially impo an in he mHeal h con ex . Some p i acy pape s ocused on mHeal h a e discussed below. A heu is ic assessmen app oach is p oposed in [23] o assess mHeal h applica ions o sel - acking. They use di e en sou ces (such as GDPR o usabili y) o build a lis o 26 heu is ics in ou ca ego ies (no ice o awa eness; choice o consen ; access o pa icipa ion; and social disclosu e usabili y). Then hey analyze 64 sel - acking applica ions o check i hey comply wi h hei heu is ics. They de elop a sco ing me hod, bu hey a e no exclusi ely based on GDPR as a sou ce o he heu is ics. On he o he hand, [24] assess p i acy isks on 79 applica ions ce i ied as sa e by he UK Na ional Heal h Sys em. Unlike o he assessmen s, hey enume a e a se ies o opics o p i acy policies, based on In o ma ion Commissione ’s O ice o Uni ed Kingdom and he Da a P i acy Ac o 1998 ( epealed nowadays). Then hey analyze i e e y opic has been add essed o no in he p i acy policy o he assessed applica ions. An analysis o he 600 mos used mHeal h applica ions as o May 2013 is p esen ed in [25]. I shows ha only 30% o he applica ions had a p i acy policy on ha da e. They also analyze he a e age leng h o he p i acy policies and assess he anspa ency o he p i acy policies based in i ems conside ed mos impo an by use s. A compa ison o diabe es and men al heal h Indian applica ions is p esen ed in [26]. They use eadabili y es s and o he eadabili y me ics (such as wo d coun and wo ds pe sen ence) o conclude he e is no signi ica i e di e ence be ween he wo ypes o applica ions. III. METHODOLOGY A. P i acy scale design Analyzing p i acy in mobile applica ions is no a simple ask. The e a e many ac o s ha a ec p i acy. Some o hem can be analyzed by audi ing he app i sel , while o he s equi e audi ing he applica ion’s p o ide ( o example, o check i i s in e nal p ocedu es comply wi h GDPR). In he one hand, audi ing he applica ion includes analyzing secu i y measu es o e i y i he applica ion con ains any known ulne abili y o checking i he eques ed da a a e necessa y o he pu pose o he applica ion. On he o he hand, audi ing he applica ion’s p o ide is ela ed wi h he concep o accoun abili y desc ibed abo e. Thus, he con olle mus implemen a secu i y policy in o de o p o ec pe sonal da a, mus deploy a p i acy policy, mus keep eco ds o p ocessing ac i i ies and, when necessa y, mus ca y ou a da a p o ec ion impac assessmen be o e p ocessing da a. In his pape we ocus on he documen ha desc ibes he p i acy policy since i is he i s poin o con ac be ween he use (known as ‘da a subjec ’ in he GDPR) and he applica ion p o ide . P i acy policy is one o he pilla s o p i acy. This way, GDPR pu s emphasis on he p i acy policy and he in o ma ion ha mus be p o ided o he da a subjec , when p ocessing pe sonal da a. Ou objec i e in his pape is o analyze any p i acy policy in a sys ema ic way so ha we can ob ain a sco e o assess he quali y o p i acy. A icle 13 o GDPR es ablishes he kind o in o ma ion ha mus be p o ided o he da a subjec i he app collec s da a om him/he . Mo eo e , pe sonal da a mus be p ocessed in a anspa en manne , acco ding o a icle 5. The e o e, i is essen ial o e i y ha in o ma ion gi en is cohe en wi h he obliga ion o anspa ency. We also conside he ecommenda ions o he Spanish supe iso y au ho i y on p i acy policies [27, 28]. This way, we de ine a scale ha builds a p i acy sco e based on some o he ea u es o he applica ion p i acy policy. The conside ed i ems and he sco e o all o hem a e he ollowing: •Iden i y o he da a con olle : 0 poin s i he iden i y is omi ed, 0.5 poin s i only pa ial in o ma ion is gi en and 1 poin i ull da a is p o ided (including name o he da a con olle , pos al add ess and elec onic add ess). •Iden i y o he ep esen a i e when he con olle is ou side he EU. 0 poin s i no ep esen a i e is iden i ied and 1 poin i i is. This i em is no applicable (N/A) i he da a con olle is inside he EU. •Con ac de ails o he DPO: 0 poin s i no in o ma ion is p o ided and 1 poin i con ac in o ma ion is p o ided. In la ge scale mHeal h applica ions, we mus be awa e o a icle 37 o GDPR, as i manda es ha a DPO mus be designa ed in his case. •Pu poses o he p ocessing: 0 poin s i no pu pose is p o ided, 0.5 poin s i he in o ma ion p o ided is oo gene ic and 1 poin i speci ic in o ma ion is gi en. •Legal basis o he p ocessing: Possible legal bases a e enume a ed in a icle 6 o GDPR. 0 p s i no legal bases a e p o ided, 1 poin i hey a e. •Legi ima e in e es s: i he p ocessing is based on legi ima e in e es om he con olle . 0 poin s i no in o ma ion is p o ided, 1 poin i i is. This i em does no apply (N/A) i he legal bases o he p ocessing a e di e en om legi ima e in e es . •The ecipien s o ca ego ies o ecipien s o he pe sonal da a: 0 poin s i he ecipien s a e no p o ided, 1 poin i hey a e. We conside ha his in o ma ion mus be gi en e en i he e a e no ecipien s o he collec ed da a. •T ans e s o pe sonal da a o a hi d coun y (ou side he Eu opean Union): 0 poin s i no in o ma ion is p o ided, 0.5 poin s i he in o ma ion p o ided is oo gene ic, and 1 poin i speci ic in o ma ion is gi en ( o example, i ecipien holds a ecognized ce i ica ion such as P i acy Shield [29]). We conside ha his in o ma ion mus be gi en e en i he e a e no ans e s. •Da a s o age pe iod: 0 poin s i no in o ma ion is p o ided, 0.5 poin s i he in o ma ion p o ided is oo gene ic and 1 poin i he pe iod is gi en o , a leas , he c i e ia o de e mine ha pe iod. •Exis ence o da a subjec ’s igh s: 0 poin s i no igh s a e poin ed ou , 0.5 poin s i he in o ma ion p o ided is oo gene ic and 1 poin i speci ic igh s and a me hod o exe cise hese igh s a e enume a ed. •Exis ence o he igh o wi hd aw consen a any ime: 0 poin s i no in o ma ion is p o ided and 1 poin i i is. This i em does no apply (N/A) i da a subjec ’s consen is no a legal base o p ocessing. •Righ o lodge a complain wi h a supe iso y au ho i y: 0 poin s i no in o ma ion is p o ided, 0.5 poin s i in o ma ion p o ided is oo gene ic and 1 poin i speci ic in o ma ion is gi en (such as he con ac o he supe iso y au ho i y). •Obliga ion o p o iding da a: whe he he p o ision o pe sonal da a is a con ac ual equi emen , o a equi emen necessa y o en e in o a con ac . This i em also conside s whe he he da a subjec is obliged o p o ide he pe sonal da a and he possible consequences o ailing o p o ide such da a. 0 poin s i no in o ma ion is p o ided, 1 poin i i is. •Exis ence o au oma ed decision-making, including p o iling: 0 poin s i no in o ma ion is p o ided, 0.5 poin i in o ma ion p o ided is oo gene ic, 1 poin i speci ic in o ma ion abou he logic used and he possible consequences o he da a subjec is p o ided. Also 1 poin i he e is no p o iling and i is men ioned. Table I shows a summa y o all he de ined i ems. We also include an i em numbe o u u e e e ences in his pape . TABLE I. ITEMS IN THE PRIVACY ASSESSMENT I em I em No. Sco e Iden i y o da a con olle 1 0: no in o; 0.5: pa ial; 1: ull Iden i y o he ep esen a i e 2 0: no in o; 1: In o p o ided; N/A: no applicable DPO de ails 3 0: no in o ; 1: In o p o ided Pu poses o he p ocessing 4 0: no in o; 0.5: gene ic; 1: speci ic Legal basis o he p ocessing 5 0: no in o; 1: In o p o ided Legi ima e in e es s om con olle 6 0: no in o; 1: in o p o ided; N/A: no applicable Recipien s (o ca ego ies) o he pe sonal da a 7 0: no in o; 1: in o p o ided In e na ional ans e s o da a 8 0: no in o; 0.5: gene ic; 1: ull de ails o no in e na ional ans e s Pe iod o which da a will be s o ed 9 0: no in o; 0.5: gene ic; 1: speci ic Exis ence o da a subjec ’s igh s 10 0: no in o; 0.5: gene ic; 1: ull Exis ence o igh o wi hd aw consen 11 0: no in o; 1: in o p o ided; N/A: no applicable Righ o lodge a complain wi h a supe iso y au ho i y 12 0: no in o; 0.5: gene ic; 1: speci ic Obliga ion o p o ide pe sonal da a 13 0: no in o; 1: in o p o ided Exis ence o au oma ed decision- making o p o iling 14 0: no in o; 0.5: gene ic; 1: speci ic o no p o illing/au oma ed decision done B. Me hods A sys ema ic sea ch was ca ied ou on 30 h Janua y 2019, in he Spanish e sion o he wo leading mobile app s o es: Apple S o e and Google Play. All ele an mHeal h apps o BC we e iden i ied using he keywo ds “b eas cance ”. Fi s ly, duplica es we e emo ed. Fo hose apps de eloped o bo h ope a ing sys ems, iOS and And oid, only And oid e sion was selec ed. Ti les and desc ip ions o he esul ing mHeal h apps we e e iewed and assessed o eligibili y agains he selec ion c i e ia. Inclusion c i e ia: •App is in ended o BC pa ien s •App is ocused exclusi ely on BC •App con ains use in o ma ion o allows use o sha e hei opinions/da a Exclusion c i e ia: •Desc ip ion is no w i en in English •P i acy policy is no w i en in English no in Spanish i a ailable •App is no ocused on BC •App is ocused on cance in gene al, al hough BC is men ioned in he s o e desc ip ion. •App is no ocused on BC sel -managemen o suppo •Apps in ended o o he s han BC pa ien s IV. RESULTS App s o es sea ches esul ed in 154 mHeal h apps bo h in Apple S o e (24.7%; n=38) and Google Play (75.3%; n=116). A e emo ing duplica es, i les and desc ip ions o 148 mHeal h apps we e assessed o eligibili y by wo esea che s. Disc epancies we e esol ed by consensus. Finally, 10 mHeal h apps me he selec ion c i e ia. One o hem was excluded due o no being eely a ailable. The emaining 9 mHeal h apps we e downloaded o assess he p i acy policy applying he p oposed scale. Ano he one was excluded a e ins alling i due o ailu e in he equi ed egis e p ocess. Table II shows he selec ed applica ions conside ing he inclusion/exclusion c i e ia. Fo con enience when analyzing he esul s, applica ions we e agged om app1 o app8. TABLE II. S ELECTED APPLICATIONS App name De elope Ope a ing Sys em Label Becca – Suppo and Guidance B eas Cance Ca e And oid App1 My B eas Cance Coach Genomic Heal h, Inc. And oid App2 B eas Cance Suppo MyHeal hTeams And oid App3 OWise B eas Cance Px Heal hCa e B.V. And oid App4 Booby app – The B eas Cance App Booby app And oid App5 B eas Cance Manage @Poin o Ca e iOS App6 B eas Cance Ally The Uni e si y o Michigan iOS App7 B eas Cance Su i o Po able Medical Technology iOS App8 To ob ain he p i acy policies o he applica ions, we i s downloaded hem using he link p o ided in he applica ion page in he co esponding applica ion s o e (Google Play o App S o e om Apple). I no link was p o ided, we looked o a p i acy link in he de elope ’s web. Then, we ins alled he applica ions in a sma phone, used i , and checked i he p i acy policy was he same as he one shown in he de elope ’s websi e. Mos o applica ions (5 ou o 8) did no p esen any p oblem wi h hei p i acy policies. Howe e , app1 p i acy policy in he applica ion s o e was di e en om he one a ailable when using he applica ion. Thus, he la e was analyzed. In addi ion, we we e no able o use app7 because i asked o a PIN om a doc o o hospi al. In his case, we ha e used he policy a ailable in he de elope ’s web page. Finally, app8 does no ha e an a ailable p i acy policy, nei he in he de elope ’s web (p i acy policy is applied o all he se ices p o ided by he de elope ) no in he app i sel . Thus, we decided no o conside his applica ion. Some bad in en ion was obse ed in some o he applica ions. App3 p i acy policy is ambiguous abou he pe sonal da a ha hey p ocess. Thei de elope s claim ha hey do no collec use -speci ic iden i ying in o ma ion, jus o say a e wa ds o whom hey do sha e pe sonal in o ma ion. Mo eo e , app5 de elope s claim ha hey do no hold any pe sonal iden i iable da a, bu hey allow s o ing pho os and audio, which migh be used o iden i y people. Besides, hey decla e ha hey comply wi h GDPR, wha is illogical wi h no p ocessing pe sonal da a. Table III shows he sco es ob ained a e assessing he p i acy policies o he 7 apps. We de ined a pe cen ual sco e. A sco e o 100 means a o al accomplishmen o GDPR. I an i em is no applicable, i is no conside ed in he inal sco e. TABLE III. P RIVACY SCORES App name Label Ope a ing Sys em Sco e Becca – Suppo and Guidance App1 And oid 57.7 My B eas Cance Coach App2 And oid 25.0 B eas Cance Suppo App3 And oid 78.6 OWise B eas Cance App4 And oid 31.8 Booby app – The B eas Cance App App5 And oid 29.2 B eas Cance Manage App6 iOS 71.4 B eas Cance Ally App7 iOS 34.6 The i s issue we no ice is he low le el o compliance wi h he GDPR in gene al. Only 3 ou o 7 p i acy policies sco e a 50% o highe . The highes sco e eaches 78.4% and he lowes one only sco es 25.0%. All p i acy policies gi e ull o pa ial in o ma ion abou he iden i y o he da a con olle and do speci y he ca ego ies o ecipien s o pe sonal da a (I ems 1 and 7). We conside ha his is posi i e aspec . In addi ion, all applica ions excep one explain he pu poses o he p ocessing (I em 4). This con as s wi h he ac ha none o he 5 applica ions whose da a con olle is ou side he EU iden i ies a ep esen a i e (I em 2). No in o ming abou he iden i y o he ep esen a i e is conside ed a se ious in ingemen unde he Spanish da a-p o ec ion law [30]. Al hough he Da a P o ec ion O ice (DPO) is compulso y when p ocessing special ca ego ies o da a a la ge scale (such as da a conce ning heal h), only 2 applica ions ha e designa ed a DPO (I em 3). P i acy policies do no in o m p ope ly abou use ’s igh s, such as he igh s o access, co ec o dele e in o ma ion (I em10). Only 3 o he applica ions gi e enough in o ma ion abou use ’s igh s, 1 o hem gi es only pa ial in o ma ion and 3 o he policies do no e en men ion hese igh s. Fu he mo e, only 3 apps men ion he da a subjec ’s igh o lodge a complain wi h a supe iso y au ho i y (I em 12). Mo eo e , hey do no de ail some aspec s, such as how o ind con ac in o ma ion o he supe iso y au ho i y. P i acy policies also su e om lack o in o ma ion ega ding he exis ence o no o p o iling echniques. Only 3 policies in o m use s abou he exis ence o p o iling, bu hey do no gi e much in o ma ion abou he p ocess, ega dless GDPR s a emen s (I em 14). Besides, 4 o he policies gi e some in o ma ion abou how long pe sonal da a will be s o ed (I em 9). 3 o he policies do no poin ou he legal basis o he p ocessing (I em 5). I em 5 is impo an , since ha ing a legal basis o he p ocessing is he i s i em ha should be accomplished, and no conside ing i is a e y nega i e aspec . Rega ding he in o ma ion abou ans e s o a hi d coun y, only 2 applica ions ully in o m use s abou he ans e s ou side he EU (I em 8). 2 o hem in o m abou he loca ion o he da a ea men bu do no desc ibe he measu es p esc ibed by GDPR. Finally, 3 applica ions do no poin ou whe e he da a a e s o ed. We did no ind any ele an conclusions o I ems 6 and 11. Table IV summa izes he esul s ha ha e been explained abo e. TABLE IV. A CCOMPLISHMENT OF GDPR ITEMS SUMMARY I em No. Full in o ma ion Pa ial in o ma ion No in o ma ion No applicable 1 5 2 0 0 2 0 0 5 2 3 2 0 5 0 4 5 1 1 0 5 4 0 3 0 6 2 0 1 4 7 7 0 0 0 8 2 2 3 0 9 2 2 3 0 10 3 1 3 0 11 3 0 1 3 12 0 3 4 0 13 1 0 6 0 14 0 3 4 0 V. CONCLUSIONS Since p i acy is one o he main conce ns o b eas cance (BC) pa ien s, i is impo an o analyze whe he mobile applica ions comply wi h he ecommenda ions o he au ho i ies. In his pape , we ha e analyzed he p i acy policy o 8 mobile applica ions o BC pa ien s o check i hey comply wi h he Gene al Da a P o ec ion Regula ion (GDPR). Wi h his aim, we designed a scale o assess some o he i ems ha mus be complied by he apps. Ou scale builds a p i acy sco e based on some o he ea u es o he applica ion p i acy policy. We de ined a pe cen ual sco e, whe e a 100-poin sco e means a o al accomplishmen o GDPR. The low le el o compliance wi h he GDPR was qui e su p ising, as only 3 ou o 7 p i acy policies eached a 50-poin sco e. As posi i e aspec s, all p i acy policies gi e ull o pa ial in o ma ion abou he da a con olle . Nea ly all o hem explain he pu poses o he p ocessing, and mos o hem speci y how long hey s o e he da a. As nega i e ea u es, only 2 apps ha e designed a Da a P o ec ion O ice , only 3 o hem in o m he use s abou all hei igh s, jus 3 o hem epo abou he exis ance o p o iling echniques, and ano he 3 do no poin ou whe e he da a a e s o ed. Legal bases a e also o en o go en. Fu he mo e, when he apps a e ou side he EU, none o hem iden i y a ep esen a i e in he EU, while only 2 ully in o m he use s abou he ans e s ou side he EU. Thus, we conclude ha he e is s ill a long way o he whole compliance o GDPR. App de elope s mus be conce ned abou his ma e . REFERENCES [1] Wo ld Heal h O ganiza ion. “Global S a us Repo on Noncommunicable Diseases 2014”, ISBN 9789241564854, 2014. [2] R.L. Siegel, K. D. Mille , and A. Jemal, “Cance s a is ics, 2016”. CA Cance J. Clin, ol. 66(1), 2016, pp. 7-30. [3] A. K. A ing on, L. Golds ein, L. K upe , C. Vi o, J. Yim, and S.L. Chen. “Li e expec ancy a e cu a i e-in en ea men o b eas cance : Impac on long- e m ollow-up ca e”. Am. Su g, ol. 80(6), 2014, pp- 604-609. [4] S.P. Leong, Z.-Z. Shen, T.-J. Liu, G. Aga wal, T. Tajima, N.-S. Paik, K.Sandelin, A. De ossis, H. Cody, and W.D. Foulkes. “Is b eas cance he same disease in Asian and Wes e n coun ies?”. Wo ld J. Su g, ol 34(10), 2010, pp. 2308-2324. [5] R. De Angelis, M. San , M. P. Coleman, e al., and EUROCARE-5 Wo king G oup, “Cance su i al in Eu ope 1999–2007 by coun y and age: esul s o EUROCARE-5—a popula ion-based s udy” Lance Oncol., ol. 15 (1), Jan. 2014, pp. 23-34. [6] J. Cho, S.-Y. Jung, J. E. Lee, E.-J. Shim, e al. “A Re iew o B eas Cance Su i o ship Issues om Su i o s’ Pe spec i es” J. B eas Cance , ol. 17 (3), Sep. 2014, p.189. [7] M.E. Hewi , A. Bamundo, R. Day, and C. Ha ey. “Pe spec i es on pos - ea - men cance ca e: quali a i e esea ch wi h su i o s, nu ses, and physicians”. J Clin Oncol, ol. 25, 2007, pp. 2270-3. [8] B. Caul ield, and S. Donnelly. “Wha is Connec ed Heal h and why will i change you p ac ice”, QJM, ol. 106(8), 2013, pp. 703-7. [9] R. Whi ake . “Issues in mHeal h: indings om key in o man in e iews”, J. Med. In e ne Res, ol. 14 (5), 2012, e129, a ailable a : h p://dx.doi.o g/10.2196/jmi .1989. Las consul ed: Feb ua y 2019 [10] W.T. Riley, D.E. Ri e a, A.A. A ienza, W. Nilsen, S.M. Allison, and R. Me mels ein. “Heal h beha io models in he age o mobile in e en ions: a e ou heo ies up o he ask?”. T ansl Beha Med, ol. 1, Ma ch 2011, pp. 53-71 [FREE Full ex ] [doi: 10.1007/s13142- 011-0021-7] [Medline: 21796270] [11] G. Giun i, D.H. Giun a, E. Guisado-Fe nandez, J.L. Bende , and L. Fe nandez-Luque. “A biopsy o B eas Cance mobile applica ions: s a e o he p ac ice e iew”. In e na ional Jou nal o Medical In o ma ics, ol. 110, 2018, pp. 1–9, a ailable a : h ps://doi.o g/10.1016/J.IJMEDINF.2017.10.022. Las consul ed: Feb ua y 2019 [12] S. M. Phillips, D. E. Con oy, S. K. Keadle, C. A. Pelleg ini, G. R. Lloyd, F. J. Penedo, and B. Sp ing. “B eas cance su i o s’ p e e ences o echnology-suppo ed exe cise in e en ions”. Suppo . ca e cance O . J. Mul ina l. Assoc. Suppo . Ca e Cance , May 2017. [13] N. H. Nguyen, N. T. Hadg a , M. M. Moo e, D. E. Rosenbe g, C. Lynch, M. M. Ree es, and B. M. Lynch. “A quali a i e e alua ion o b eas cance su i o s’ accep ance o and p e e ences o consume wea able echnology ac i i y acke s”, Suppo . Ca e Cance , ol. 25 (11), 2017, pp. 3375–3384. [14] N. Ribei o, L. Mo ei a, A. Ba os, A.M. Almeida, and F. San os- Sil a. “Guidelines o a cance p e en ion sma phone: A mixed- me hods s udy”, In e na ional Jou nal o medical In o ma ics, ol. 94 (1), Oc obe 2016, pp. 134-142. [15] M. C. Robe son, E. Tsai, E. J. Lyons, S. S ini asan, M. C. Swa z, M. L. Baum, and K. M. Basen-Engquis , “Mobile Heal h Physical Ac i i y In e en ion P e e ences in Cance Su i o s: A Quali a i e S udy,” JMIR mHeal h uHeal h, ol. 5, no. 1, p. e3, 2017. [16] G. Giun i, J. Kool, O. Ri e a Rome o, and E. Do onzo o Zubie e. “Explo ing he Speci ic Needs o Pe sons wi h Mul iple Scle osis o mHeal h Solu ions o Physical Ac i i y: Mixed-Me hods S udy”, JMIR mHeal h uHeal h, ol. 6 (2), Feb. 2018, e37. [17] S. R. S oyano , L. Hides, D. J. Ka anagh, O. Zelenko, D. Tjond onego o, and M. Mani. “Mobile app a ing scale: a new ool o assessing he quali y o heal h mobile apps.”, JMIR mHeal h uHeal h, ol. 3 (1), Ma . 2015, e27. [18] Eu opean Pa liamen and Council, “Regula ion (EU) 2016/679 o he Eu opean Pa liamen and o he Council o 27 Ap il 2016 on he p o ec ion o na u al pe sons wi h ega d o he p ocessing o pe sonal da a and on he ee mo emen o such da a, and epealing Di ec i e 95/46/EC (Gene al Da a P o ec ion Regula ion)”, EUR-Lex, a ailable a : h ps://eu -lex.eu opa.eu/eli/ eg/2016/679/2016-05-04. Las consul ed: Feb ua y 2019. [19] G. Con issa, K. Doc e , F. Lagioia, M. Lippi, H.-W. Mickli z, P. Palka, G. Sa o , and P. To oni. “Au oma ed p ocessing o p i acy policies unde he EU gene al da a p o ec ion egula ion”. 31s In e na ional Con e ence on Legal Knowledge and In o ma ion Sys ems, JURIX 2018, He Kas eelG oningen, Ne he lands, Volume 313, pp. 51-60, doi: 10.3233/978-1-61499-935-5-51. [20] G. Fox, C. Tonge, T. Lynn, and J. Mooney. “Communica ing compliance: De eloping a GDPR p i acy label”. 24 h Ame icas Con e ence on In o ma ion Sys ems 2018: Digi al Dis up ion, AMCIS 2018, Hya Regency New O leans, Uni ed S a es. [21] W.B. Tes ay, P. Ho mann, T. Nakamu a, S. Kiyomo o, and J. Se na, “P i acyguide: Towa ds an implemen a ion o he EU GDPR on in e ne p i acy policy e alua ion”, IWSPA 2018 - P oceedings o he 4 h ACM In e na ional Wo kshop on Secu i y and P i acy Analy ics, 2018, pp. 15-21, Tempe, Uni ed S a es, doi: 10.1145/3180445.3180447 [22] K. Renaud, and L.A. Shephe d. “How o make p i acy policies bo h GDPR-complian and usable”, 2018 In e na ional Con e ence on Cybe Si ua ional Awa eness, Da a Analy ics and Assessmen , 2018, A icle numbe 85514422018, Glasgow; Uni ed Kingdom, doi: 10.1109/Cybe SA.2018.8551442. [23] L. Hu on, BA. P ice, R. Kelly, C. McCo mick, AK. Banda a, T. Ha zakis, M. Meadows, and B. Nuseibeh, “Assessing he P i acy o mHeal h Apps o Sel -T acking: Heu is ic E alua ion App oach”, JMIR Mheal h Uheal h 2018;6(10):e185, doi: 10.2196/mheal h.9217. [24] K. Huck ale, J.T. P ie o, M. Tilney, P-J. Benghozi, and J. Ca . “Unadd essed p i acy isks in acc edi ed heal h and wellness apps: a c oss-sec ional sys ema ic assessmen ”, BMC Medicine201513:214, doi: 10.1186/s12916-015-0444-y. [25] A. Sunyae , T. Dehling, P.L. Taylo , KD. Mandl, “A ailabili y and quali y o mobile heal h app p i acy policies”, J Am Med In o m Assoc. 2015 Ap ;22(e1):e28-33, doi: 10.1136/amiajnl-2013-002605. [26] A. Powell, P. Singh, and J. To ous, “The Complexi y o Men al Heal h App P i acy Policies: A Po en ial Ba ie o P i acy”, JMIR Mheal h Uheal h 2018;6(7):e158, doi: 10.2196/mheal h.9871. [27] Spanish da a p o ec ion agency, “In o me sob e polí icas de p i acidad en In e ne . Adap ación al RGPD”, a ailable a : h ps://www.aepd.es/media/es udios/in o me-poli icas-de-p i acidad- adap acion-RGPD.pd . Las upda ed: Sep embe 2018. Las consul ed: Feb ua y 2019. [28] Spanish da a p o ec ion agency, “Decálogo pa a la adap ación al RGPD de las polí icas de p i acidad en In e ne ”, a ailable a : h ps://www.aepd.es/media/es udios/decalogo-poli icas-de- p i acidad-adap acion-RGPD.pd . Las upda ed: Sep embe 2018. Las consul ed: Feb ua y 2019. [29] Eu opean Comission, “Commission Implemen ing Decision (EU) 2016/1250 o 12 July 2016 pu suan o Di ec i e 95/46/EC o he Eu opean Pa liamen and o he Council on he adequacy o he p o ec ion p o ided by he EU-U.S. P i acy Shield (no i ied unde documen C(2016) 4176)”, a ailable a : h ps://eu - lex.eu opa.eu/eli/dec_impl/2016/1250/oj. Las upda ed: July 2016. Las consul ed: Feb ua y 2019. [30] Spanish O icial S a e Gaze e, “Ley o gánica de p o ección de da os y ga an ía de de echos digi ales”, a ailable a : h ps://www.boe.es/eli/es/lo/2018/12/05/3. Las upda ed: Sep embe 2018. Las consul ed: Feb ua y 2019.