i@ II ill ,,@l t iltit* 0 1 9 * AUTOMATED IMULTIPIIASIC 14EALTII TESTING AND TliE REGIONAL MEDICAL PROGRAMS A Report of a Subcommittee of the NATIONAL ADVISORY COUNCIL ON REGIONAL MEDICAL PROGRAMS May 11, 1971 IZockvil.'].e, li@,3rylatid 11, 1.971 To: THE NATIONAL ADVISORY COUNCIL 0,Nj REGIONAL MEDICAL PROGRAMS In November, 1970 Council requested a subconinittee examination of automated ,,,-"-tnultiphasic health testing as a regional medical program activity. The subcommittee ttiat,received this assignment is submitting its report and recommeiidatio ns. The projects that regional medical programs have funded have not operated long enough to accumulate experience that can be evaluated fairly. Summaries of their direct cost funding and their statements of purpose and benefit are included in the report. Most of the RMP funded projects emphasize early detection of disease and preventive medicine as a principal reason for their existence. Health education and conservation of physician time are also stressed as purposes of these projects. I Vllth these arguments in mind, the subcommittee conducted a conference on multiphasic health testing in Bloomfield Hill.s, Michigan on April 29-30. At the conference the subcommittee heard expert presentations on the states of the arts of: diagnostic testing; secondary prevention of chronic disease; utilization of technicians and automation in testing program; and practical experience with the acquisition and application of personal health data in a variety of episodic and preventive care programs. The subcommittee has concluded that regional medical programs should withhold funding from any new multiphasic health testing projects, including those that have been approved, but were not funded by the end of 1.970. The subcommittee is convinced that intensive efforts should be made to gather and evaluate the experience that will be gained in the projects alreadN, funded. The subcommittee also believes that modelling the natural histories of chronic diseases through systems analysis techniques should be explored. Such explorations might identify diseases for which preventive care of predictable value can be designed and identify the gaps in our knowledge of others. Michael J* Brennan, M.D., Chairman John E. Kralewski, Ph.D. Alexander M. McPhedran, M.D. Clark H. Millikan, M.D. AliTO',T"'TL@D i'i'iT,,rT.P 11-@S-"C III-Al,'ril Tr,@" 11@G AI;D TliE REGICI'@@L liEDICAT, PROGR/i@-IS Multiphasic health testing, as examined by the subcommittee, is the .1 application to a defined population of a uniform battery of tests capable of-detecting disease or high risk of disease in persons in whom these conditions have not been recognized. Some multipliasic heal.th testing systems employ several batteries of tests and measurements, each of which is applied uniformly to a segment of the test population that is distinguished by specified demographic, medical history, or other characteristics. Multiphasic health testing may or may not employ automated or computer-aosisted. means of acquiring medical histories, obtaining measurements and storing and retrieving information. Automated Multiphasic Health Testing Among purists in the field, the term "automated" is reserved for those systems in which automation is most extensive. In these systems not only many individual operations, but the flow of examinees through the testing establishment and much of the interpretation of the data obtained are computer- co nttqlled, The costs of.equipping and installing such a system can be amortized within a reasonable length 6f time only by a high volume of service or a high fee for service. In general the extent to which a healtli testing system is automated is determined by the volume of service it is designed to perform. Services Provided by Multiphasic Health Testing Multiphasic health testing systems acquire data about examinees which are s@imil-ar in range of application to those acquired by a physician administering a health check-up with the aid of laboratory services, The multip!iasic testing systems employ non-physician technicians to acquire'and process the data for physicians. This relieves physicians of routine data collection and of the more mechanistic interpretations. Hopefully, this reduces the time tile ph sician spends with each y patient, and enables him to provide his higher levels of professional service to more patients. Patients also save time in the testing process by comparison with the tiirlc- required for an individualized physician directed cherk-tip. One trip through the multiphasi6 testing process any replace separate visits to the physician's office, a clinical laboratory and an x-ray department. 2 - Both physicians and patients li,-Avc, recoSyni.z(@C,@ those be tiefits accruing from the consistent structuring of the m,,ilti.pl-i,3,-ic tc@.@;ti.ng and reporting. Automated systems are credited with additional benefits in improved reliability. It is claimed that automated test performance, data recording and reporting are less error prone than similar operations controlled by hurnans, The uniformity of terminology and format of a multiphasic system's reports offer another time saving advantage. Any physician familiar with the system can utilize any of its reports, whether he is at the initial point of referral or comes into the case later. In-Syster.i Economic Considerations Multiphasic health testing systems may be classified as either integrated, or free-standing. For purposes of this report, integrated systems are those that serve fixed groups of physicians and their patients. The physicians may be in solo practice associated with a specified hospital or hospitals or clinic on a conventional fee for service basis, or they may be meiiil-)ers of one or more groups providing health and medical care under a prepayment plan. Integrated testing systems are owned, under exclusive contract to, or otherwise control-led by,.the medical care groups they serve. Fre6-standing testing systems accept examinees on aelf-referral, physician referral or under contractual arrangements with labor, employer, or other organizations. Their services are rendered caveat emptor, and the physicians or patients they serve can exercise only indirect controls on their operations through economic action. All multiphasic systems inevitably face economic problems. In addition integrated and freestanding systems each have special economic problems. The Volume-Capatity Equ6tion A crucially important determinant of the survival of a multiph6sic testing system is the unit cost of its service. In each testing system's environment, there is a maximum.cost which cannot be exceeded for long, and there is a minimum cost below which the service cannot be provided. The limits of this range will be more sharply marked for free-standing than for integrated systems, but both must accommodate to very reallimits. The relationship between the volume of service performed and the intended design capacity plays an important role in determining t-he unit cost of the service. The cost of maintaining the system in readiness t.e,,-ids to be close to the cost of It f-@ll I C@-,pac@[,L:@'. the volume of service falls si.g-.L-ii.ficantlv @)elcw --,ztp,,@city, the cost of e,-cli unit of service tends to mount above that in the design. IIlicTi Cic,.,,iarid for service exceeds the capacity of the svsteir,, excessive requests must be denied or back logged, or extraordinatry operations must be undertaken. A decline in volume mav be more readily tolerated in an integrated system, where health care demand and efficiencies generated by the testing may justify some subsidy of its operations. Demand in excess of capacity may be attractive to the freestanding testing system which may even gain.prestige by developing a waiting list. In the interest of production neith-er type of testing system can long tolerate a wide disparity between design capacity and actual volume of service. Patient Satisfaction Another economic constraint is imposed on the unit costs of both types of systems by the need to satisfy the examinees. In any mul.ti.p asic testing system it is necessary to process a large number of patients whose tests will .yield negative or normal reports. if the testing program is beneficial to the individual, the exam-Lnee should continue to utilize it periodically, because his condition may be expected eventually to change. Continued participation of the "normal" examinee is important to the health of the system also, to maintain its volume of service at or near design capacity. The examinees interest in periodic testing is directly dependent upon his perception of the benefits of the exercise in comparison with his outlay of time and money. Thus, a health testincy system that is-dependent upon periodic examination C3 of the same population is constrained to control its changes and to utilize the opportunities provided by the system for health education of the examinee. Health Care Demand in the Integrated Testing System One of the purposes of multiphas.ic testing.in the integrated system is generation of increased demand for the diagnosis and treatment of pre-symptomatic, early, or previously unrecognized chronic-disease. Early treatment of such disease C> is alleged to improve the chance of restoring health and to reduce the lifetime health care needs of the patient,'if not by curin his condition at least by 9 reducing his needs for more costly forms of care. . To real.ize-the benefits visualized by this hypothesis, the-patient must actively seek and persist in the indicated treatment or management of his disease. Equally important to sticcessfttl-reqliz,-.tion of the hypothesis is the capacity of the system to provide the indicated treatment and long term management. The benefits of trading participation i.ii low cost ambulatory care or future high cost I,-te-stage care accrue to patients i,71io r)crsi-,t in, and receive effective early stage care. To achieve this benefit for its patients therefore, the integrated care system must respond to the early treatment needs it discloses with services that are effective both in control of disease and in retention of the patient's cooperation. In the early years when screening is applied to a closed population, the ambulatory care workload may increase rapidly, whil.e the saving in more tostly forms of care is not realized. In a closed population limited capacity system.. the revelation of additional patient needs tends to frustrate their fulfillment, because it overloads treatment resources. The medical care services are pressed to expand and increase in cost, whil.e cost control pressures tend to reduce the long term gains for patients who stay in the system, to reduce patient satisfaction, and stimulate patient defection. In this kind of situation the svstem is,pressed economically toward: (a) abandonment or rigid control of i-ts response to the preventive care indications of its multiphasic screenings, and (b) toward dilution of the ill and high risk components of its patient population by acqui.ri,i-Lg new lo@i-risk persons or by reducing its accessibility to high risk persons. The integrated m@ultiphasic testing service that provides pre-admission or routine work-up services for the staff of a hospital or clinic is less affected by these economic pressures. Its services generally are performed as adjuncts to care of acute disease and the attending physicians are not barred from active participation in the promotion and implementation of preventive care programs. The free standin g heal.th testing svstem may be almost entirely free of these pressures. Both types of testing systems can prosper at relatively constant levels of operation as long as their servi.ces fulfi.],l their limited promises. Intrinsic Values of Multiphasi.c Health Testin, All of the foregoing economic and functional views of multipl@iasic health testing have been stated without reference to the intrinsic health care values of testing, 5 - in the long run the individual examined or pst-ici@t I)iys the dollar costs his perL--o-,-,al rcsc)tlrccs, of the testing by d)-@,wiiig either @ii oi-ic- @7ay or anc@',: or by drawing on protection offered to hini-by government. The patient also invests time and emotion in the testing. The patient may receive three kinds of return on his investments. -Returns in Emotional Reassurance The processes of obtaining professional health care are costly in'money, time and emotion.- Any increase in any of these types of expenditures wil.1 be resisted unless it is associated with perception of offsetting reductions in the others. To most patients, multiphasic health testing initially offers appreciable emotional reassurance. Its promise of earl.y detection and preventive care seems both effective and up to date. Itis lists of facts about the patient assure him that his physicians will not overlook and he himself will not orget to mention important features of his health status. The pitch of the whole activity encourages a hopeful attitude toward.his heal.th. For all patients and particularly for those inclined to pre-occupation with health,, (the wcl-ried well) these potential returns are f.regil.c. They are subject to abrupt reversal by insensitive behavior of medical care personnel. and by errors or breakdowns in the testing, reporting, and medical care responses of the system. Returns in Process Efficienc)r Almost all multi.phas]'Lc health testing systems, whether designed primarily to facilitate preventive care to control disease, or to assist in work-UPS 0 acute illness are wholl.y or partially additions to the patient's health care costs. To be justified, these new expenditures should yield new benefits, If multiphasi6 testing of patients with acute 'Complaints reduces hospital.i.zat on costs, or time spent in visits to laboratories, or shortens the diagnostic process in almost any way, the economic return on the investment can lie made obvious by a few words from the attending' physician, Multiphasic testing for preventive care and disease control is less obviously profitable. The ultimate economic return, if any accrues, is so long in coming that the patient is constrained to rely upon his conviction.s to justify his expenditures. 6 - in the detached view, the possibility of lo'@ll term is very fragile because it depends upon adec.,uat-@ peT-f,@)ritian(-,e of i.,i@,ilt.Lpl.e, srnall, complexly reacted acts over a long period of time. Proof of ultimate gain in life cost for medical services through preventive care is also apt to be elusive for most chronic diseases. Partly because the prevention or containment processes applicable to much of chronic disease are so lengthy and subject to so many,kinds of lapses and ruptures, evidence that reductions in their lifetime costs in time or money are actually feasible will be difficult to develop. It fol.lows that evaluation of AiN@qlT as a factor in preventive medicine is a long and complex task, Returns'in Hebltli Improvement Actual gains in longevity or even in health status resulting from multiphasic health testing are very difficult to measure for any chronic diseases. Much of this difficulty is rooted in lack of controls and in the lack of knowledge of the natural. histories of the diseases themselves. For some diseases we do not know how many or which of the cases detected or forcast would riot progress to impairfunction, or threaten life, even if left undisturbed, For others it is ai yet-ipipossibl.e to predict the effectiveness of available control measures. For others the risks associated with available treatment have not been fully evaluated. Similarly, many of the parameters utilized in intiltiphasic health t6st3.ng are not completely understood. This increases the difficulty of measuring effectiveness of disease control, because it casts doubt upon the validity of the initial detection. clearly an investment in multiphosic health testing as a predictable way to obtain long-term health status maintenance or improvement remains highly speculative. Current Status of @'iultiphasic Health Testing iti-Re@i.onal. Medical Programs egi.onal Medical Programs have invested in twelve projetts that feature Ten R multiphasic health testing that is ni-itomated tolsorne degree. The regions represent the east and west coasts, and southeastern, Midwestern and mountain states. The' populations to be served represent inner city ghetto residents, rural disadvantaged is groups, employee groups and cross-sec.tions of hospital admissions. 7 - One of the projects is primarily tiittire, At7if)iig the oti-iers, the i.r-nc?cliat@e. objectives most frequertl.,.@, Trenti..olled (See Tables 2 and 3) are education of the examinees and of physicians and the .3xi.ofiiatic preventive services of screening and casefinding. The broader purposes cited include demonstration of the A2,IHT clinical approach, introduction to medical care of inadequately served groups and modelling of improved dollar, manpower and medical efficiency for the health service system. We have little information on plans for continuation of these projects after their grant periods, (Table 4) Some of the projects sponsor single examining establishments. Some have or,plan to open several. One is concentrating on mobile service in a medically disadvantaged area and one concentrates on disadvantaged school children. Each project has some degree of automation, and each b8s its own plan of tests and measurements. Apparently none of these project -e plan automation as extensive as that developed by Kaiser Permanents, the leader in this field. At this time we have little experience by which to evaluate these projects. Six of the twelve projects received their initial awards in 1970, two in 1969, two in 1968 and the developmental project began in 1.967. It has taken +these screening projects from nine months to two years after their initial awards to attain operational status resembling that visualized in their initial, plans. In the more comp@ex systems, de-bugging of the mechanism, building up the volume of business, and activating follow-up of examinees take more nine. Most of the projects have declared the intent to study themselves, Patient and physician acceptance of the system, fol.lo@i-up of examinees with abnormal testss system efficienc and cost-benefit comparisons are incl.uded in the plans, y as are the significance of the tests and various epidemiological topics. Unfortunately, the studies planned to date will be of an internal nature, The sponsors do not appear to have the base-line date and control capabilities that would enable them to compare their AMIT systems with alternative systems for achieving the disired purposes. Even under the best of circumstances, it wil,l be difficult 'to get highly rel.iable system comparisons within the relatively brief spen of a three to five year RMP grant. Current study plans of the project no not include evaluation of their effects on regionil.iza tion of health care services 3,970 (c'c-o T@)blcl 1) for the The tota'. of ti i r c) i't I @ di.r(@ct costs of these projects w,-Is ,5,920,000. The everigt- di.rect cost for a project ),c--r about @260,001,), 'I'heso with attendant overhead costs, riia,ce ul) a si@,vii'ic,3nt sharij of tlic, total amounts of risk ., A- money available to al.1 regional medical, programs. or to any one regional medical. program. The 1970 total for direct costs of $2,884,500 al.one i-s 6.7% of the $43,246,000 amount awarded for project grants in that year For this reason as well as because of the lack of consensus on so many of the AMHT hy otbeses, the que stions before the council remain insistent. p Is it a concern of the li@IP system to make the definitive evaluations of the AMIIT hypotheses? Can the Regional Medical Programs (or any other single subsystem of the medical economy) complete these evaluations in time and with sufficient authority to steer the mtion's (or a regionts) application of the valid hypotheses? Will the hypotheses be tested in the open mArl,,et, regardless of regional medical program intervention?. To this sub-co-,nittee it is clear that automated inultiphasic heal.th testing is 6 mode of obtaining patient health data that is being tested in many of the applications for which such data are required. Disregarding the many live y questions that present themselves about the validity of the tests employed and of tile purposes to which the data are applied, automated multiphasic health testing can be an efficient method of acquiring and processing the data. Three conditions must exist in any situation in which these systems are to become 6fficient. First, there must be a sufficient volume of testing to keep the cost of operating the system commensurate with the efficiencies achieved, Second, the automated riultiphasic-testing system must be linked with equally efficient systems for feeding examinees into the testing process and following through on its data output, to accomplish the purposes for which the data are acquired. The testi ng system is not an end in itself, unless its sol.e objective is acquisition of epidemiologic information, which is a question of questionable relevance to the M4P mission. 9.- T I, 1. @u L ed he a ItIt te@(3 t: @,-nf, -i 1 IC II 1 Sr,@@i t IIOD tll,'3 t@IIle, only when the choice is supp(,3rtcd by ttiorotic-,Ii CO,@;t-I)C'Tlf!fit cc)-,iipirisons iqi.th alternative me thods of data acquisition Bec,-,iise of tl,,e i.,iri,,e investment involved an autorritted testing r,@rsterii in @iTi economically marg-.Lral situation can become a monster, whose needs can distort the health care system it is intended to serve. The sub-committee has not found evidence that AtiliT itself has directly improved regional deployment or utilization of beal.th service resources. Some of the purposes served by A@filT methods may have such effects but AMIIT in itself can only contribute efficiencies or inefficiencies to the performance of the tasks assigned to it. The subcommittee believes that regional medical. programs should concentrate on improving deployment and utilization of practi.ce-reidy health care systems o predictable effects. So manv of the hypotheses on-which A@IIIT rests are unproven that it cannot be credited with predictability as positive influence in health care. The sub-committee also finds that the uncertainties surrounding preventive health care are not likel.y to be resolvable within the life span of a regional medical program project, Recommendations of the Subcommittee The-subcomm@ittee proposes that the Council issue the following recommendation: "At this time eleven regional medical programs -have funded projects that feature automated w-ultiphasic health testing. The purposes of these projects present a-fair representation of the purposes for which patient health status data are acquired. "Abtomated health testing is very costly. The influe nce of the projects in which it appears on regional. deployment and utilization of health care services is highly unpredictable, For these reasons Council recommends that no new projects featuring automated health testing be funded. "The Council further recommends that the Director, RIIPS, and the appropriate regional niedi.cal programs, coordinating with the National Center for Health Services Research and Development, Community Health Services, the Nati.onal Center for Health Statistics, the National. Institute of General Medical Sciences and other interested agencies, institute consultation and investigation to: grant oper,,.-)t,ioiis of proAects currently into irj!P and other pro-ccts @L i-ie 2. Build i e j,,ita, Foils and measures of progress for cohorts of persons i,71-iore initial ritiltirgliesic tests were positive, negative and refused, such populations as urban and rural poor, employee groups, hospital. and clinic pitients, to help resolve debate about tl-ie affects of ititil.tiphasic testing on qual.ity of and access to health care services and the regional deployment and utilization of health care resources- an@ 3. Utilize ems analys .s and all available epidemiol.ogic information to stimulate atu@al histories of diseases and identify those for which' secondary prevention might be feasible and acceptable in cost." F-4 4'.- w 4-> co cl CL tD > 0 0 n) 0) E, (D 0 rD 0 rt l@. pi 0 r) cri 010 m H. 0 (Li (D H- rb o 0 D) (D Hi HI c :3 0 :-I 0 0 ro :i CL rA 0 CA En 0 11 :L) :3 2) CD r_ t- CL (D rt 05 0) (D 03 ro co pi H. 0 (D :3 0 rt (D o rt li co 0% 00 00 011 0 0 (D rt 0 fD -4 fD Ul r) rt C) 0 f-3 0 rt 0 tt) Li %D0 C) C)O li Ln ri Lo H. rt H (D rt w co p- co C) Ln LI> 00 Ql %D0 LI) co Ln Ln c: D3 0 00 IQ LI) LI) ch P- t-h t- oo ON t- Ul co C:> -4 C:> C) co -i (Z) ON I_n VI ON U.) C) 00 Ln C) 0% co %D i-n LTI w cr, co -i co LI) rt t- co Ca L.)i LIS (7% U) 00 C:) -i 00 w t,i ib0 N) t- 0- - P- L.> Lo "" I ri)m 0 .rt Na ci -4 t-n t-n kD 0 (D 03 a) 0) rr 0 ri) :-I rD H. H. (4 rD H- Pt Hi ii) 0 0 :-3 0 0 CL (A 0) c: ;j E* E(i 0 En CT m 0) rD cn 0 rt tQ IQ OD -i 00 m rt %,O 00 -4 1.0 kD %D co F-4 n tt t') En m En C) CL 1-h t'J En r-t s 0) t-i 03 0 Cl. r_ lb m 0) 0) C.L CD m 0 CD " (r) ti C. rt El .4 tn U) r_ Ell 06) rt rt rt D) 4 m < (D f)0 H. (D m m (1) r-r 2) (D rt 0 (D m (D n)5 I- I Ca (D rt 0 m < :3 ri) 0 :7 rt fl) H. 0 rT, 0 rt D) rT rt H. t. rt (?) 10 0 H. rt :3 om El 0) (D 0 1-d H. 4 OrQ OQ :3 (D H. CL w < Z (D v r_ 0 (D (D rn :i 0 rD A) (D F ::r . r_ DQ :i 0 0 m 0 fl) cq rt :3 CD 0 tA 0 co P. r.0 H. CP 11 m m !Z ti " rn rt (D ri) rt :3 M rt (D tzi N (D rt m m rD i,- 03 CL H. aQ C) 10 (D CA @ i ti 0 0 " m (D m vi 0 cn cn Pi 0 (:) tA rn cr 0 H. 03 0) tD rt C-t a) (D 03 pi 0 co 03 rt co C13 11. t. rt, m 0 " 0 0 (D (D r- M 10 :3 0 a). IL -3 H. CL 0 OQ U) r-r m H. cri (D 0 m .0 f- :i (D 2) rt rt (D :3 (D 0. oq cn 0 coli (D F4 (D ;j m C4 0 :3 DI ri (D pi rD -130 pi 10 11 rt (D :3 CL 0 < :3 Fl 1-h rt (D :3, 0) 0 lb 0 03 rt :3 CL w (D rlt m- m (ID a 0 :3 co CL. 0 H. c: 0 F- CL rt < 0) (D C6 o rt 0 tD (b (D 0) 0 0 :3 rt m " :3 0 :3 t-I 9 0 rt CL ao 0:3 03 H. 03 0 m (D 0 rD D) ti :3 rt 0. Et) rt 0) OIQ (D m rD 0 03 CL CD ri a) < m 10 H. Di rt <0) m (1) 0 0 0 :1 C3 rt :3 la 0 0) O' rt r_ - rt :I (IQ 0 c0 10 F- (D rp 0) :J F- 0) :3 r- O. < I pi 03 0) a, 0 m r? 0 0 rr 0 0 E3 CL c H. pi E3 rlt rt rt o10 0 0 0 El la 0 0 03 rt 0 m 03 0 o 0) 03 0) pi m :i oq < (A rt rT ti m rt rr 0 0 01,0 m 0 0 Fl- CL 0 pi :i c: 0 rt 03 H- :4 m @ Cl) I w (D rD tD m H- CD < 0 ID ro :3 0 r4 @ rt 0 H. co 0 ti F4 V4 U) (D 0 0 0 t'D m El :x :n ri) < rD rD rt 0 (D H. 03 :3 0 a) 0 rt t:i ti It (D m m m (D rD rD 0 m 0 0 :o 0 0 @O :3 co rD :3 :i :I :3 :3 :i ri) 0 0 En co U) co co rn 0 En co r? rt rt rlt la rlt rt rt ri- Pi pi pi pi El pi -A m (b Di 03 0) D3 m Ca co m rt rt rr I r? rt 0 0 r3 0 0 1-n 00 LA 03 t3 w rt pi 0 9) CrQ m 0 01 0) H. pi H. (D 0 0 0 0 :3 CL m c: ;j 0 03 fi) w H. ft CL CL m rt 03 ra ID Ch a) 0 U) -i 00 00 fD 0 Ul w ct) o rD a) 0 rt 0 m 0 m CL tD CL Pi co ti pi rt P. El :3 CL II r. k- c: 0 :J< la 0 (D < 0 la 0 Li. p- to > M @ ti :i t- H. a) En Ct) (D 14 (a rt H- 0 otD m rn0 m N (b rA 0 rt rt (13 H. co H. rt CL (D CL r? 0) zr ct (D PI H. rt CL0 OQ rt Z r- CL (D r?lm " 0 0 0 -(D 0 (D Ca 0 0 tl- rl' rt 0 H. Z a) W. rt rlo rt rt cn :r 0 ti) (Ao m 0 co rt 0 (D 0 (D (D 03 0) U) 00 CL CLM pi to H. rt t- M,:J Cu< P- < F- Fl- 0) (D En (D 0 (D rA H. CD CT m CL (D (D co M co o rt (.) rA ri) r_ pi 0 0 0. m pi pi ib :3 co 0 F4 pi lb m CT < 0 rD 0co (:r rrM 0 tD CL 0 H. 0 m 0 ?I t- 0 -, 0 0 rD < =1 0 m e, H. :.g n) ct CQ o) 1-h ib (D (D 0 a) (t) (D 0 D) oq 0 10 t,- (D I.. (D :3 -4 CL 0 =- m rt CL rt :i 23 r?, ti 0 ct CL rt rt m I= 0 . " m . -0 m m m 0 14 ra 03 :3 C) OQ Di 0 0 0 z co :r 10 :3 U) I'. DQ" pi rt 03 co 0) 0 pi GQ M 03 rt M :3 (D aq (D F4 rt rD o ;3 OQ0 :3" 0) :$ rt CL m < (D CL el rD tb 0 z L4) tu CL 0 (D C.4 co 0 CL 0rt 03 H. (D0 0 ft co D) Fl. 0 0. ul 0 mtt ::r 0 CL :i rD 0 pi :3 m (D H 0 ib P. 0 zr :4 ::r m 0) " " z C6 rt5 m pi Ca 0 rt 03 < " 0(D 10 CD (D w 0 t. :3 rD H. lb0 C6 pi. lb (4 H. :3 0 Co H. 0 " lb I-. rt th to rt D3 o r?, cn m m (,) a3 (D ?I 0 14 (D CL CD 0) m =r 0 Cl) t- ft =r. th 14 fD O ti (D (D rD M- rn E3 Pi Oo CL) (D CL m rh M) rt C3 F. ti m co :i0 pi [- 0 m H. ib cn co :3 (r) 0 :3 rt rt t-I rt :3 - :I 0 O!Qm n rT pi P- 0 U) rt lb. ?I 0 9 (D t.. rt CL. P" ;j 0 rA rt 10 < :3 0 m 0 H. (JQ m 0 03 F-4 td n 0) td ti F3 m PI :3 0 0 14 0 m 10 rt H. rt r_ 0 rA CT a) ti rt :1 0 P- la rlr m 0 tD oq CL m m m co co 0 CL rt CL 0 m CD pi crq CT CL n CL a) < (D H. 03 ro E3 0) pi r) Pi cn 0 0 m 0) 0 rt co CL - I oq DI (D < 00 pi rD 02 (b < CL :3 cl r? (D m 0 OQ :3 c: CL c: 0 0 r t" D3 la I... rT rt 0 0 0 10 (D 0 co rt 0 rt 0 Ul Ul :r 00 0 rD rt td z > rD CL 0 :3 0 0 fD t%j ffi F4 rT (D :i rT rt (D ra rt > t'D P- 0 F- rt t- rD o 0 (D pi m eD (D CL rA rt CL cr 0 rA i-n 4 0 0 0 0 :r 0 0 n) rD rt k,' 0 :3 rA :1 N I rD o ti 0 CA (D H. PIT m FJ- (D o 0 F9 0 0 < m 0 a, rA z :3 E0 0 ti rt c :i fb :3 ft 0 C)% co -i 00 00 W t7-1 z 0 0 L- m 0 rt 0 CL F- (D Ul (D C) C:> 0 P- 0 C) 0 C:) C> .0 o 0 rt co a-Q C:> 0 m rt 1-1 0 (D Ca C/) rt rr C: C6 ?,- 'O rt IV :3 10 0 A 0 ct 0 .PIT PI. cn 0 m rt 0 o Cl) 0 0 co :i =, rD A) rt 0 (b CD m m ul 03 0 0 0 m (D ;3 rD ri 0- ri pi m En n Ct) ;j 0) E* Ca co LO ti 0 (D (D tl z Ln 03 (D rA co rt CD CL 4A 0 4A H. Li Ln ti L'i H. Ili c: 03 o OC) 0 . 0 0 0 rt pi . m co 0 m Ln 0 F- C) 0 0 9 0 0 0 03 0 DI C:) 0 C) C-) 0 (D rt 03 0 0 0 tn C) ft :3 rA 0 rt rt r) rA 0 pi C> :3 fD t- C: 0 m P- rt pi M ;I 9 rlt m 0) m :I la 0 rt m CL C: ti m rt 03 m 0 ai tzj En t:i m 00 t) 03 0 0 M Ln 0 CD 0 0 0 0 C) 0 la 0 la 0 " t%) 03 0 10 la rlt m rA > CT 0 :3 rt pi (D rA " rt M . rt co Pi 0 CD . Pi C:) :r o 0) < " rt t-d 0 m 0 :3 0 :3 :I rt 03 0 Ci) 0 H. aq kA 0 Pi 0 co rA CA :1 m CT ?- PI z I* rt 10 ' rT :i 0) . H. LI) m m ::r 03 ril H. el m rt a) < (D 03 pi :3 CL Pi m ct m m (D pi (L CA :3 :3 :i co m r_ 10 r_ (A ti m < U) a] (D 0 rlt m 03 m rt 11 rA 9 2) < ti z Pi m U) pi m 0 r_ rt :3 m ro 0 (D c: 0) rt m o m eD m m 0 C) x 0 C) C) H. 0 C) m cr m m Cl) (D (D Cl.