R e s e a r c h A r t i c l e A n A n a l y s i s o f t h e C o m p l e t i o n o f P h y s i o t h e r a p y C l in ic a l R e c o r d s in G a u t e n g P r o v i n c e . ABSTRACT: The purpose o f this study was to rate the completion o f physiotherapy assessment, treatment and progress records in the Gauteng province o f South Africa. There is a dearth o f literature on physiotherapy patient records, y e t it has been demonstrated that clinical records have significant implications f o r quality o f care, resource allocation, research and f o r professional litigation. A combined retrospective and prospective research design, using a quality assurance instrument was used to rate the completeness o f physio­ therapy records obtained fro m multiple study sites. Breakdown in clinical recording in the follow ing areas is described: • areas o f care (private and public sectors), • patient conditions (e.g. orthopaedic and surgical) • patient categories (in and out patients), • section o f the record (e.g. demographics, physical examination), and • providers o f care (physiotherapists, physiotherapy assistants and physiotherapy students). The 644 records analyzed scored a mean overall com pletion rate o f 64%. There was a significant difference observed in overall record com pletion (p = 0.0004) between private and public sector providers. Significant differences were also observed f o r each section o f the record. Factors most associated with a high degree o f completion o f physiotherapy p a tie n t records included; private sector physiotherapy services, use o f pre-form atted assessm ent charts and c lin ic ia n s’ participation in relevant continuing education. Physiotherapy patient records in Gauteng were fo u n d to be incomplete. The implications o f this finding in an increasingly competitive global, national, corporate and domestic healthcare arena include; physiotherapy service quality, professional litigation, resource allocation and policy development, as well as professional growth, develop­ m ent and satisfaction with ones ’ career. Recomm endations on how the physiotherapy profession can improve clinical recording fro m an educational, clinical and managerial perspective are suggested. M'KUMBUZI VRP, BSc Honours in Physiotherapy, UZ’; EALES CJ, PhD2; STEWART A , MSc Medicine3 1 S ta ff D e v e lo p m e n t F ellow , U n iversity o f Z im b a b w e , Faculty o f M e d ic in e , R e h a b ilita tio n D e p a rtm e n t. H e a d , School o f T h e rap e u tic Sciences, Faculty o f H e a lth Sciences, U n iversity o f th e W itw a te rs r a n d . S en io r Lecturer P h y s io th e ra p y D e p a rtm e n t, School o f T h e rap e u tic Sciences, Faculty o f H e a lth Sciences, U n iversity o f th e W itw o te rs r a n d . K E YW O RD S: COMPLETION, CLINICAL PH YSIO TH ERAPY RECORDS. This research article is an extract of part of a research report submitted to the Faculty of Health Sciences, University of the W itwatersrand, Johannesburg, in partial fulfillment of the requirements for the degree M aster of Science in Physiotherapy, (August 2 0 0 1 ). INTRODUCTION Background P hysio th erap y is both a hum ane en d eav o u r and a business operation (Bauer 1989). In the case o f modern day physiotherapy practice, which includes C O RRESPO N D EN C E TO: VRP M ’kum buzi University o f Zimbabwe, Faculty o f M edicine, Rehabilitation D epartm ent P.O. Box A 178, Avondale, Harare Zim babwe Tel: +263-4-791631 o r +263-91-317086 Fax: +263-4-791995 or 724912 Email: rehab@ healthnet.zw or vy vpiw ai2000 @hotmail .com com m unity and primary levels of care and private as well as public sector services, physiotherapy involves colla­ boration with a num ber o f stakeholders. These are ministries o f Health, Education, Welfare and Social Services, non- govern­ mental organizations (N G O ’s), medical aid com panies, the clien ts’ family and caregiver as well as the individual client himself. Physiotherapists are therefore accountable to their clients, to society and to the funding source. This “ wide range o f accountability indicates a wider range o f monitoring indicators in the w hole process o f co n tin u o u s and ongoing evaluation o f any rehabilitation service” (M yezw a 1997). In an endeavour to transform the health system in the USA, H aughom (2000) proposed that the solution to the health care crisis lay in taking a lesson from industry and providing a higher quality product more efficiently. The w ay to accom plish this goal is to “collect better inform ation” , and apply it to improve care (H aughom 2000). Aim of Study The aim o f this study was to determ ine the degree o f com pleteness o f physio­ therapy patient assessm ent, treatm ent and progress records. M otivation and Importance of Study This study presum es that physiotherapy patient records are important. However, 18 S A J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) mailto:rehab@healthnet.zw there is a dearth o f literature relating to inform ation on physiotherapy patient records. In spite o f this situation, hospital records are sent to central statistical and planning offices to review health sector p erform ance and determ ine resource allocation. Shrinking health budgets at global, national, corporate and household levels necessitate an evaluation o f clinical work. The ultim ate purpose o f a patient record is to provide the data so that clinical caregivers can make timely and accurate decisions. It is thus generally accepted that an analysis o f patient records is one way o f dem onstrating workload as well as the quality o f service. Quality is important in the evaluation o f physiotherapy practice, as a profes­ sional, organizational, administrative and econom ic enterprise. Methodology P ilot and F easibility1 Studies A pilot study was conducted to deter­ mine a reasonable null hypotheses test level for the main study; re-fam iliarize the researcher with the adm inistration o f the QA instrument and validate the demo­ graphic questionnaire data instrument. A three-part feasibility study was conducted to determ ine the relative proportions o f physiotherapy clinicians pro v id in g physiotherapy services in selected institutions in a given month and the monthly patient loads. Further, other study variables such as the nature o f the physiotherapy record for in and out patients which could affect physio­ therapy record completion were investi­ gated in order to design a study procedure for sampling and retrieval. Conclusion and Recommendations of the Pilot and Feasibility Studies. Study Design The pilot and feasibility studies led to the conclusion that an analysis o f physio­ therapy clin ical records in G auteng province using a retrospective m etho­ dology was at the time not entirely feasible. This conclusion was based on the prem ise that a record analysis would have to take into account the following: • The different levels of physiotherapy care • The different physiotherapy service providers, and • T he use o f a scientifically sound study design. The com bined prospective and retro­ spective design was deem ed necessary in order to include multiple study sites with sim ilar and dissim ilar health infor­ mation systems, and thus gauge more accurately the record keeping practices across the province. Secondly research assistants were used to: • Blind the principal investigator to the identity, training level and w ork ex p erien ce o f the physiotherapy provider who had written the record, prior to and during the process o f rat­ ing each physiotherapy record. • To speed up and make more efficient the data collection process. Study Objectives 1. To determ ine the degree o f com ple­ tion o f physiotherapy patient records in G auteng Province 2. To com pare the degree o f completion o f p h y siotherapy p atient records between those written by public sec­ to r p h y sio th erap y prov id ers and records w ritten by priv ate sector physiotherapy providers. 3. To determ ine sections o f physio ­ therapy records that are complete, and incom plete 4. To determine patient factors associated with a high degree o f com pletion o f physiotherapy records 5. To determ ine assessm en t chart characteristics associated with a high degree o f co m pletion o f p h y sio ­ therapy records. 6. To determ ine physiotherapy service providers’ characteristics associated with a high degree o f com pletion o f physiotherapy clinical records. Study Sites A sam ple o f seven physiotherapy service providers was selected in the G auteng province. Two hospitals represented public sec­ tor physiotherapy providers and were selected because • They are tertiary levels o f care, • Employ physiotherapists and physio­ therapy assistants, • As teaching hospitals they are used for physiotherapy students’ clinical practice and • F o llow ing the feasib ility study a method o f retrieving in-patient and out-patient records could be designed Four private physiotherapy practices were random ly selected from providers listed in the G auteng private practi­ tioners’ telephone directory and a fifth practice was a convenient selection, because it had becom e increasingly difficult to obtain consent from private p ractitioners. F urther, the feasibility study had shown the need to use m ul­ tiple sites with sim ilar and dissim ilar health information systems in order to gauge more accurately the record keep­ ing practices across the province. Physiotherapy Record Selection A sample size calculation determ ined the need to obtain an overall sam ple o f at least 369 records, and 260 records each from the public sector and private sector, in order to detect a 10% differ­ ence in com pleteness at the 0.05 level o f significance. All records for patients presenting to physiotherapy in the selected institu­ tions for the first tim e for the current problem during the month o f February 2001 were included in the study sample. In this way, the overall sample yielded would be proportional to the volume o f new patients presenting to physio­ therapy in each institution, and from each clinical unit and area. The month of February was a convenient selection as it gave sufficient time to audit trail and com plete the study within the first half o f the year. M ain Study Procedure R esearch assistants conducted the first part o f the study. This part included: • Selection o f physiotherapy records using the selection criterion. The 1) A separate report on the m ethods and results o f the fe asibility study will be published subsequently. SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 19 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) selection criterion was physiotherapy records o f patien ts who received physiotherapy for the first time for the current problem during February 2001. • Labeling the service provider as a physiotherapist, physiotherapy assis­ tant or physiotherapy student. • Recording the dem ographic details o f service providers responsible for writing selected patient records • Recording patient characteristics of patien ts w hose record had been selected, and • Recording the chart characteristics o f selected patient records. The second part o f the study was to rate the com pletion o f physiotherapy records. The principal investigator did this exclusively to elim inate inter-rater variation. V arious techniques w ere used to select all records m eeting the inclusion criteria, the most im portant technique was the com bination o f the prospective and retrospective design in the public sector hospitals, audit trail and peer back tracks. T he secretary -recep tio n ist in each practice retrieved patient records. The principal researcher was blinded to the profiles o f the provider in two ways: • R ecords w ere first rated, then the physiotherapy providers’ details obtained from the secretary, • The researcher knew none o f the physiotherapy providers. Instrum entation The Q uality A ssurance Instrum ent This QA instrum ent was used to rate the com pleteness o f physiotherapy assess­ ment, treatm ent and progress records. The QA instrum ent was developed in the M idlands province in Zim babw e between 1994 and 1999 ( M yezw a et al, 2001). The salient features o f the QA instrum ent are percentage ratings on the degree o f com pletion o f sections o f patient records as shown in Table 1. The QA instrum ent was piloted in May 2000 on 35 physiotherapy records from Hospital 2 and adapted for use in this study without modification. The Dem ographic D ata Instrum ent The d em ographic d ata included the characteristics o f service providers, such as gender, age, years o f training, years o f experience and place o f w ork. The instrum ent also elicited data on the condition o f patients assessed, and the character o f the assessm ent charts used. T hese variables have been found to influence chart completion in the lite­ rature. The dem ographic data chart had also been piloted and a few m odifica­ tions made. D ata A nalysis D escrip tiv e statistics w ere used to summarize the dem ographic characteris­ tics o f the phy sio th erap y clinicians resp o n sib le for w riting the p atient records sampled and the score and rating of the overall com pleteness o f physio­ therapy records A one sample “t” test was calculated to test the null hypotheses for the over­ all com pletion o f physiotherapy patient records. A tw o-sam ple “t” test was calculated to test the nul I hypotheses for the overall com pletion o f physiotherapy records written by private sector providers com ­ pared to those written by public sector providers. Pearsons’ correlation test was calcu­ lated to determ ine the level o f correlation between the percentage physiotherapy record com pletion and a) The percentage rating for the physical exam ination section o f the record (including its objectivity). b) A ge o f the physiotherapy service provider c) Num ber o f years o f physiotherapy training and, d) Number of years of working experience. Secondary Analytical Statistics The above statistical analyses were done for a total sample o f 644 (694 less 50 illegible m icrofiche records) physiothe­ rapy patient records. O f these, 266 were written by private physiotherapists and public sector physiotherapy providers wrote 428. Since the sam ple o f records (n = 644) had been written by only 46 physiotherapy providers, (31 public sector providers and 15 private sector providers), the analysis o f 644 records could be interpreted as records written by 644 independent clinicians. However, it was likely that w here multiple records w ritten by one clin ician had been included in the sam ple for analysis, these records may in actual fact repre­ sent the rating o f only one record. Thus, a second database was g en er­ ated where each physiotherapy provider was represented only once by his or her mean. All third year and fourth year physiotherapy students were grouped to represent one provider each respectively. This was done for the following reasons: • Students often share the management o f a p a tie n t, and • It was not always possible to identify the student.. The one sam ple “t” test for overall com pletion and the tw o-sam ple “t” test to com pare record com pletion between the two sectors o f physiotherapy care w ere recalcu lated using the second database. Power: The sam ple size o f 31 and 15 was used for public sector and private sector pro v id ers respectively. Both sectors had a power in excess o f 95% (public sector 100%, and private sector 99.99% ) in detecting the observed dif­ feren ces for phy sio th erap y clinical record com pletion. A ccu racy: T he 95% confidence limits for the overall com pleteness of physiotherapy records were calculated to be within 4.4% of the observed mean. RESULTS Study Sites O ne university teaching hospital (1100 beds in use), a non-teaching general hospital (780 beds in use), two big private p h y sio th erap y p ractices and three small physiotherapy practices in Gauteng province were the study sites selected. Table 1: Percentage Ratings o f the Q A Measuring Instrument. Section of the record Demographic data History social, past and present medical Physical Examination, findings and problem list Treatment Aims and Plan Progress Note, discharge and referral summary % Rating 15 15 50 10 10 2 0 SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) Table 2: Study Sample. Institution Hospital 1 Hospital 2 Private practice 1 Private practice 2 Private practice 3 Private practice 4 Private practice 5 Total Number of Records obtained 279 151 127 55 24 6 52 694 Figure 1: Patient Conditions o f Selected Records.Physiotherapy Clinical Records The num ber and source o f physiothe­ rapy clinical records analysed is shown in Table 2. A total o f 694 physiotherapy records were retrieved from the selected insti­ tutions (430 from public hospitals and 264 from private providers). Fifty micro­ fiche film records were excluded from the statistical analysis for completion because the film was poor and physio­ therapy records illegible. The results are therefore based on the analysis o f 644 physiotherapy records.. Physiotherapy Providers Forty-six physiotherapy clinicians wrote the records included in the study; 39 were written by physiotherapists, eight (8) by physiotherapy assistants and 21 were written by third or fourth year physio­ therapy students. M ost (n = 19) o f the providers were physiotherapists based in Hospital 1, followed by students from Hospital 2. There were no physiotherapy assistan ts o r physiotherapy students providing services in any o f the selected private physiotherapy institutions. In addition 83 records (12%) were written by a male clinician and 611 (88%) were written by a fem ale clinician. The mean age o f all physiotherapy service providers was 30.3 years (SD±9.5). Two hundred and sixty-six (38.3%) were records written by private sector physio th erap ists w hile 428 (61.7% ) were written by public sector physio­ therapy providers. Level of Training Various categories o f training levels were observed. These ranged from third and fourth year students to two year trained physiotherapy assistants and physiotherapists with postgraduate qua­ lifications and or several continuing education courses. M ost physiotherapy records analyzed were written by physio­ therapists with three or four years basic training only. Duration of Physiotherapy W ork Experience Various duration o f physiotherapy w ork­ ing experience were observed. M ost physiotherapy records that were sam ­ pled were written by clinicians with over five years working experience, followed by new ly qualified clinicians and physiotherapy students with less than one year working experience. Characteristics O f Patients Whose Record Was Selected Two patient categories were used to classify patients whose record had been selected, thus 352 (50.7% ) were records o f in-patients while 342 (49.3%) were records o f outpatients. The distribution o f patient conditions o f selected records is shown in Figure 1. M ost physiotherapy patient records selected belonged to patients who had been treated for orthopaedic impairments or dysfunction. Clinical Physiotherapy Record Formats and Chart Types All (100% ) p h y siotherapy records selected and analyzed were m anually charted i.e. in a written form. In addition 530 (76.4%) were analyzed from a paper format, while 164 (23.6%) were analyzed from a m icrofiche film format. The proportions o f the chart types used for physiotherapy records were as follows: Preformatted chart = 44%, non-prefor- matted chart = 37% and partially prefor­ matted chart = 1 9 % . 2) In m ost parts o f the w orld c u rre n t physiotherapy training takes four y ears to com plete. The three y ear d egree or d ip lo m a w as done by the older ph ysiotherapists, very few o f w hom w ere in cluded in the study sam ple from the private sector. SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 21 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) Table 3a: Record Completion fo r the Total Study Sample. Public Sector Record Completion Private Sector Record Completion O verall Record Completion 54% (SD±15.2)3 78.4% (SD±23.5) 64% (SD±23.7) N = 378 N = 266 N = 644 RECORD COMPLETION Completion of Physiotherapy Clinical Records The mean overall com pletion rate for the 644 physiotherapy records is shown in Table 3a. The com pletion rate for the private sector (n = 266), and public sector physiotherapy records (n = 378) is also shown in Table 3a. T he public and private sector mean com pletion rate rem ained significantly different with the private sector scoring a higher com pletion rate, (p = 0.0006). Record Completion for Physiotherapists in the Public and Private Sectors T he private physiotherapy practices in this study em ployed physiotherapists The differences in the mean com ple­ tion between public and private sector p hysiotherapists rem ained significant (p = 0.0004). T he co m pletion o f p h y siotherapy records was also significantly different betw een p h y siotherapists (66% ) and p h y sio th erap y assistan ts (43% ) and The public and private sector mean com pletion rate was significantly dif­ ferent, with the private sector scoring a higher com pletion rate (p = 0.0006). The recalculated mean com pletion for physiotherapy record com pletion w here each provider is represented by his or her own mean (n = 46) is shown in Table 3b. only. Thus a tw o-sam ple “t” test was calculated to com pare clinical record com pletion betw een p h y siotherapists only from the public sector and physio­ therapists in the private sector. The result shown in Table 3c is based on an analysis that excluded records written by physiotherapy assistants and physio­ therapy students in the public sector. between physiotherapy students (63.83%) and p h y sio th erap y assistants (43% ) (p<0.0001). Record Completion and Related Completion Determinants The following associations (Pearsons’ correlation test) were calculated: • com pletion o f physiotherapy records and the phy sio th erap y p ro v id e rs’ age, a w eak neg ativ e correlatio n (r = - 0.2382), p = 0.1683 was calcu­ lated, indicating that the relationship was not significant. • com pletion o f physiotherapy records and physiotherapy providers’ years o f training, a weak positive correla­ tion (r = 0.2356), p = 0.1731 was calculated, indicating that the rela­ tionship was not significant and • com pletion of physiotherapy records and physiotherapy providers’ years of working experience, a weak negative correlation (r = - 0.2785), p = 0.1052 was calculated, indicating that the relationship was not significant. Record Completion for Different Chart Types Physiotherapy mean record com pletion for the three chart types was observed to be: Preformatted chart = 77.3%, (SD±16.6), partially pre-form atted 61.4% (SD+19.8) and the u n fo rm atted ch art 46.3% (SD±22.9). The mean percentage comple­ tions were significantly different between the three chart types (p = 0.0005). Record Completion for Each Patient Category R ecord com p letio n fo r in and o u t­ patients is shown in Figure 2. (Refer to next page) F igure 2: R ecord C om p letion for Patient Categories. T he difference in com pletion for in and outpatient records was significant (p< 0.0001), at the 0.05 level o f signi­ ficance, where outpatient records scored higher com pletion rates. Record Completion for Selected Patient Conditions Percentage record com pletion for differ­ ent p atien t cond itio n s is show n in Figure 3. (Refer to next page) F igure 3: R ecord C om p letion for D ifferent Patient Conditions • The difference in record com pletion was significant between all condi­ tions shown in Fig 3 (p< 0.0001). Table 3b: Record Completion fo r Sample Expressed as One Mean fo r Each Provider. Public Sector Providers' Record Completion Private Sector Providers' Record Completion O verall Record Completion 55% (SD±12.6) 78.7% (SD±6.2) 63.4% (SD±15.3) N = 31 N = 15 N = 46 (p = 0.0006) Table 3c: Record Completion fo r Physiotherapists Only. Public Sector Record Completion by Physiotherapist O nly Private Sector Record Completion by Physiotherapist O nly O verall Record Completion by Physiotherapist O nly 53.9% (SD±23.2) 78.4% (SD±15.2) 66% (SD±23.2) N = 275 N = 266 N = 541 (p = 0.0004) 3 ) L arge v alues w ere observ ed fo r the standard deviation (SD ) see D iscussion, and fo o tn o te 5. 2 2 SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) Record Completion for Each Section of the Record C om pletion for each section o f the record is shown in Figure 4. (R efer to next page) F igu re 4: R ecord C om p letion for Each Section o f the Record (W here all values shown in Figure 4 are expressed as % completion.) The treatm ent planning section o f physiotherapy records was found to have the lowest com pletion rate in both the private and public sector, as well as for the overall sample. B oth priv ate and public sector ph y siotherapy providers scored the highest com pletion rate on the dem o­ graphic section o f the clinical record. Further, in the overall sample (n = 644) that was analyzed, 481 (69.3% ) records had a hospital o r patient identity num ­ ber, while 213 (30.7% ) did not have a hospital or patient identity number. O verall Clinical Record Completion and Completion of the Physical Examination Section of Physiotherapy Clinical Records There was a high positive correlation (r = 0.9) between physiotherapy record co m p le tio n and c o m p le tio n o f the p h y sical e x am in atio n (in c lu d in g its objectivity), (p = 0.0004). DISCUSSION Physiotherapy Clinical Record Completion This is the first known study, which rates the com pletion and objectivity o f physiotherapy patient records. A m ean com pletion o f 64% 4 (S D ± 23.7)5 was found for the p h y ­ siotherapy assessm ent treatm ent and progress record in G auteng province. However, a distinct dichotom y exists for the overall com pletion o f records between public sector (54%) and private sector providers (78.4%). This differ­ ence is also evident for each section of the record. The factors that seem to facilitate a higher completion o f physiotherapy records written by private practitioners include: • Use o f a pre-form atted assessm ent chart. • A ugm entation o f basic physiotherapy training with continuing education (sh o rt courses). In particular, the O rthopaedic M anual Techniques (OM T)6 short courses which were done by most private practitioners. Two factors make this course extre­ m ely relevant; firstly, the course Figure 2: Record Completion fo r Patient Categories. Box a n d W h is k e r Plot 100 8 0 6 0 - 8 4 0 2 0 - In-Patient CATEGORY 6 4 4 cases 5 0 missing cases O ut-Patient p <0.0001, COMP-PER = Completion percentage. Figure 3: Record Completion fo r Different Patient Conditions. 80' 2 0 - Box and W hisker Plot O rtho Paeds Neuro Surgical CONDmON 6 42 cases 5 0 missing cases p <0.0001, COMP-PER = Completion percentage. Medical ICU 4 ) A second statistical analysis, calcu la te d w ith e ach p ro v id er rep resen ted o nce by his or her m ean (n = 46; o b se rv ed m ean c o m pletion = 63.38% ), su b sta n tia ted the first analysis w here each record rep resen ted its e lf (n = 644; observ ed m ean com p le tio n = 64% ). T he tw o m eans w ere not significantly different. T h is o b servation is a ttributed to the fact that the variation fo r record c o m pletion w ithin each p ro v id er w as large. T h is m eant that the d e p endence o f record co m p le tio n o b servations w as not o f significant concern. Secondly, the large sam ple size is b eliev ed to have o v e r­ c om e the d e sign e ffect o f m ultiple records se lected from e ach provider. 5 ) T he high standard d e v iation (SD ) is in d icative o f the w ide variation in c o m pletion scores observed. It is attrib u te d to the m any facto rs w hich in flu e n ce record com pletion. T hese range from p atient c h aracteristics, chart c h aracteristics, individual w ork ethic and departm ental policy and m anagem ent guidelines. 6 ) T he private pro v id ers sam pled all have significant O M T bias b ased practices. SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 23 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) Figure 4: Record Completion (or Each Section of the Recors _2> a. £ o<_> Demography History PhysicalExamination Treatment Plan Progress Note ■ Private 92.5 76.7 75.6 67.4 84 ■ Public 76.7 67.9 45.4 44.8 48.4 □ Total 83.2 71.5 58 54 63 Section of the Record (Where all values shown in Figure 4 are expressed as % completed) em phasizes ap p ro p riate d o cu m en ­ tation style, and secondly, over 80% o f priv ate p ra c titio n e rs’ patients present with orthopaedic complaints. The relevance is dem onstrated by the high com pletion o f 79.2% observed for orthopaedic patient records writ­ ten by private sector physiotherapists. • The dem ographic data section o f the record which was filled in by the patient on the first visit. • The progress note (SOAP) was well written and this too, seems to be attributable to the OM T course. It is suspected that the high com ple­ tion rates observed for the dem ographic and progress note sections were driven by the need to support appropriate billing systems, as this business aspect is critical for private practice survival. This reasoning was also put forward by the Medical Records Institute (1999) as one o f the major m anagem ent and adm inistrative factors driving the need to im prove clin ical docum entation. Furthermore, patient billing in the private sector directly determ ines individual therapist’s remuneration. This incentive does not exist in the public sector as therapists here earn a fixed salary. W hile it was observed that one o f the public sector study sites had instituted the use o f time sheets, these were not yet being used to have any bearing on staff salaries. T he w eakest section o f clinical records written by private practitioners w as the treatm en t planning section (67.4%). Many practitioners defended this weakness in this section o f physio­ therapy clinical records by saying that the process occurs mentally. However, this defense was not acceptable because when different therapists are required to treat the patient, they are not privy to this mental processing when it has not been docum ented - thereby raising ques­ tions as to the continuity, efficacy and cost - effectiveness o f physiotherapy care. Further, if an activity is not recorded it is reasonable to assume that it has not been carried out, rather than to assum e that the process has occurred mentally. A nother weak area observed for the private sector, was the social history aspect o f the history section. Som e therapists felt it was not relevant to docum ent in particular the smoking or drinking habits o f their predom inantly orthopaedic patients. W hatever the case, the history section needs to be relevant as well as com prehensive. The influence o f physical, political, social and eco­ nomic environm ents on the aetiology and processes o f disease and disability is w ell-docum ented (Phillips et al 1992; Tarlov 1992). The need to note these risks or agg rav atin g factors in the m anagem ent and prom otion o f health is also recognized as necessary if com pre­ hensive care is the service mission. The need to improve the psychosocial com ponent o f patient care in physio­ therapy has also been docum ented (M ’kum buzi and Eales 2001). O ne may like to argue that private physiotherapy practices also achieved a higher com pletion in the writing o f their patient records b ecau se the records obtained from this care area were written exclusively by physiotherapists, i.e. do not employ physiotherapy assistants or train physiotherapy students. However, when record com pletion was compared between public and private sector physio­ therapists (only) a significant difference (p = 0.0004) was observed between the public sector physiotherapists (53.9%) and the private sector physiotherapists (78.4%). Also, no significant difference in the com pletion score between records written by all physiotherapists public and priv ate sector physio th erap ists com bined (66%), to those written by physiotherapy students (63.8%). Finally one may also speculate that independent private ph y sio th erap ists’ records are more com plete as they need to obtain as much information as possible because, there is more often than not, no back up team work on site from other medical professionals to facilitate clinical decision making. With regards to records written by public sector physiotherapy clinicians, the dem ographic and history sections o f clinical records (although low er than the private sector) achieved a high com ­ pletion of 76.7% and 67.7% respectively. However, clerks write the dem ographic section o f patient records and the history is written by the adm itting doctor on 2 4 SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) the ward or referring doctor from out­ patients’ clinics. Thus, this achievem ent cannot be attributed to the recording done by physiotherapy clinicians. All other sections of physiotherapy patient records written by public sector clinicians scored consisten tly below 50%. The overall com pletion o f 54% in the public sector is com parable to the finding of 55% by M yezwa et al (2001), for public sector rehabilitation records from the M idlands province of Zimbabwe, prior to implementation o f a QA programme. Intensive Care U nit physiotherapy patient records were obtained exclusively from public sector hospitals. A mean completion score o f 38.5% was observed and was the worst score o f all the inde­ pendent variables. This was in spite of having a specific space to enter physio­ therapy records on the ICU chart. In m any instances the nursing sister was asked to record the statement “Physiothe­ rapy done.” Intensive care unit patients often present with co-morbidity, and it is for this reason that one might expect a m ore analytical assessm ent or at least the use o f the problem oriented medical record, in order to prioritize patients’ problem s and manage the complicated patient better. Out-patient records were found to be m ore com plete, contrary to the study by Barrie and Marsh (1992), who found a higher completion for in-patient records. This finding in this study is attributed to the fact that private physiotherapy providers who in any case scored a higher level of record completion from an out-patient population o f over 80%, co ntributed, m ost o f the ou t-p atien t records analyzed. Implications of the Research Findings The incom pleteness o f physiotherapy patient records observed in G auteng p rovince raise serious concerns. Primarily these are the quality focus o f patient care, but also the cost o f physio­ therapy care for physiotherapy funding agents. P oor p atien t records may necessitate repetition and duplication of assessm ent and treatm ent procedures. These are costly in terms o f therapist- hours, therapeutic material and financial resources, travel costs for the patient and loss of work time and production capa­ city for the working patient and his employer. The use o f physiotherapy clinical records from the Gauteng province for resource allocation, research, physiothe­ rapy service review and defense in the case of professional litigation should be done with extreme caution given the criti­ cal shortcomings in record completion. The implication of poorer records found in the public sector relative to the private sector is two-fold: • Private physiotherapy providers are not obliged to subm it their statistics and patients’ profiles to the Gauteng health departm ent, and in fact do not. Thus w hile their records are more complete and the physical examination more objective and more complete, they are not used for policy form ula­ tion with regards to health priorities for resource allocation, staff training or other plans for physiotherapy ser­ vice provision. It would seem that incom plete physiotherapy hospital data, upon which monthly depart­ mental statistics are com piled, would be the only source o f information available for use by the Gauteng health department. • The hospital environm ent is probably the most im portant interface between the p h y siotherapy profession and other medical professions, and from a quantitative perspective between the physiotherapy profession and the public. The display o f poor records which may translate to poor patient outcomes poses a serious threat to the respect, grow th, developm ent and contribution o f physiotherapy to healthcare, in one or all of the follow ­ ing ways: - Diminished patient satisfaction with the process and outcom e o f therapy - D im inished recognition o f the contribution o f physiotherapy to the wider context o f health care - D im inished satisfaction with the profession by the physiotherapy professionals themselves. Recommendations A quality assurance (QA) programme needs to be implem ented in all physio­ therapy d epartm ents. T he design o f QA must at least include the following aspects: • Establish im plicit and explicit criteria for measuring quality. Som e o f the selected c riteria may necessitate reviewing current departm ental poli­ cies and guidelines. In particular to ensure that these do not contradict physiotherapy education principles or the medico-legal obligations and expectations o f health professionals. • Design structures that specify entities such as organization, power relations and coordination o f the departm ent and, in particular specification of work procedures. • D efine the processes regarding details o f the w riting o f p atient records and their incorporation into a health information system. An area of concern with physiotherapy records is to ensure that the progress note is recorded on a daily basis, and to develop mechanisms to check that this is adhered to. Further to this, methods o f retaining raw data after the statistics have been com piled must be developed. • This study has not dem onstrated the link betw een process and outcomes. The literature illustrates that it is difficult to establish this link, (Kond- ziolka et al 1989 and Ho et al 1999). However, as beneficence is related to quality (Larabee et al 2001), it is reasonable to expect an inclusion o f the relationship between process and outcom e in any physiotherapy QA programme. This broad approach to QA takes into account that criteria, structure, process and outcom es are interrelated, and at various levels each one can affect the com pletion and therefore quality o f physiotherapy record docum entation. Therefore QA should in all areas, iden­ tify where standards are required; set protocols to set standards; or audit prac­ tices where standards exist. The use of peer chart reviews or a review o f records and prim arily basing ju d g e m e n t o f quality on the im plicit criteria of expert clinicians are participatory methods that have been used and are recom m ended in the literature (M yezw a et al 2001). T he pre-form atted chart has been shown in the literature to facilitate chart SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 25 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) com pletion (Buckley et al, 1999 and Ho et al, 1999). This too has been the find­ ing in this study. It is recom m ended that pre-form atted charts be used for physio­ therapy records as m uch as possible. Pre-form atted charts exist in m any clini­ cal disciplines as well as in physiothe­ rapy. Further, where pre-form atted charts can be designed for the major conditions seen, this should be done so that all headings are relevant, and so that speci­ fic condition considerations are included. This was observed to be in practice in two o f the private practices included in this study, where assessm ent charts were in use for spinal, peripheral, and elbow and shoulder problems. The process of developing these charts will perhaps be arduous, but a com prehensive format should be adapted, as om ission o f head­ ings will affect chart completion. S eparate phy sio th erap y clinical records should be used for in-patients in addition to recording in in-patients’ bed letters. Often in-patients are discharged from hospital prior to discharge from physiotherapy. The separate physiothe­ rapy records should be used as a co n ­ tinuation record to avoid repetition and duplication of physiotherapy procedures. With time, a clinical data dictionary of terms for com m unicating physiotherapy processes across m edical professions should gradually develop, so as to avoid the global use o f terms like “m obilize” in describing interventions done to the hand, back or gait. The participation in continuing education for physiotherapists is now mandatory for continued registration in some countries including South Africa and Zim babwe. These courses should encom pass an emphasis on appropriate docum entation, and the use o f the pro­ blem oriented medical record should be encouraged as a good m ethod to facili­ tate chart objectivity and completion. Further, clinicians should be encouraged to focus their con tin u in g education attendance to areas o f need as defined by their patient populations. This study has shown to some extent that poor com pletion o f physiotherapy records is not a result of lack o f training on proper docum entation practices. It seems, the greatest challenge is to find the determ inants o f clinically relevant behaviours in physiotherapy, and to m anipulate these determ inants to bring about the desired change in professional behaviour. There needs to be an uncom ­ prom ising declaration o f professional norms in the standards o f physiotherapy assessm ent, treatm en t and progress records. A central authority with public and private sector representatives and with sufficient powers to enforce stan­ dards is required. These include methods o f ensuring data quality such as, record system rules and patient record guide­ lines, uniform ity and consistency in c lassification system s, standard data sets, and a clin ical vocabulary for national and international com parability o f data. The professional associations together with the G auteng health depart­ ment seem to be best placed to fulfill this role. It may seem m ilitant to expect the physiotherapy professional body to take such a stance, but in view of the im plications o f poor com pletion of physiotherapy patient records for the profession, for physiotherapy care and for healthcare cost this is an immediately necessary intervention. W hile none o f the physiotherapy institutions visited were using IT (at the tim e o f data co llectio n ), this study recom m ends the initiation o f IT use in physiotherapy record systems. M any of the advantages w hich ran g e from im proving retrieval, record legibility, decreased recording time and facilitating co m pletion have been d escrib ed at length in the literature. W hile initial costs m ay be overw helm ing, only service providers who harness a com petitive edge by dem onstrating the substantial benefit to the public will survive. To succeed, the resourcing o f physiotherapy record keeping systems (incorporated with medical records), including com ­ puter networks to raise the quality of care by facilitating smart clinical deci­ sions, has to be a concerted effort on the part o f “federal, provincial and national governm ents in conjunction with the private sector” . (Fitzm aurice 1994). This will accelerate health care standards” (Fitzm aurice 1994). Physiotherapy departm ents need to use the patient records they w rite in assessing and planning for physiothe­ rapy services. Only in this way can they test the usefulness o f their practices and system s. In ad d ition, p h y siotherapy providers them selves can devise and test strategies o f care that achieve the greatest im provem ents in physiotherapy care at the lowest cost. W hen armed with such inform ation, physiotherapy can resist the pressure to accept lower levels o f quality because o f an expressed inability to afford, whether this pressure comes from the provincial health budget or m anaged health care financiers in the private or public setting. If the profes­ sion cannot resist this pressure it can at least show a truer picture o f the losses and gains. CONCLUSION P h y sio th erap y reco rd s in G auteng province were found to be incomplete, both for the overall com pletion as well as for each section o f the records. This finding com prom ises the use of physio­ therapy patient records for resource allo­ cation, research, physiotherapy service review and defense in cases o f profes­ sional litigation. Record com pletion was significantly b etter am ong p riv ate p h y siotherapy providers, however, there is room for m uch im provem ent in both sectors. Poor clinical recording of the physiotherapy assessm ent, treatm en t and progress note were found to be most evident in the follow ing areas: • Public sector physiotherapy service • ICU and surgical patients • In -p atien t phy sio th erap y records taken from public sector physiothe­ rapy providers • The treatm ent planning section o f physiotherapy records for both private and public secto r p h y siotherapy providers • The progress note section o f physio­ therapy records for public sector physiotherapy providers and • The physical exam ination sections of p hy sio th erap y clinical records, in particular those written by physio­ therapy assistants. This array o f problem s points to a th ree-p ro n g ed app ro ach req u ired to improve the quality and usefulness of physiotherapy patient records. T hese are: Y 2 6 SA J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) • The implementation o f record sys­ tems, standards and regulatory con­ trols through a broad QA programme, • The inclusion o f adm inistration, m anagem ent and clinical recording in continuing health education and professio n al develo p m en t courses and conferences • The collective and individual pursuit o f professional behaviour that is cli­ n ically relevant, and unco m p ro ­ mising o f the norms and standards of the total practice o f physiotherapy. T he com m itm ent and integrity o f many stakeholders is required to suc­ cessfully achieve these goals. These may include provincial governm ents, physiotherapy funders, physiotherapy training institutions; health information and medical records staff, physiotherapy managers and physiotherapy role models and physiotherapy professional associa­ tions. M ost important to note however, is that individual therapists need to take proactive steps to improve physiothe­ rapy clinical records. H ealth service providers including physiotherapists, have a moral, profes­ sional, societal and legal obligation to ensure that patient records are o f a high quality. This obligation is captured in the em phasis on the im portance o f p rocesses rath er than the technical practice “in the accountable delivery o f a holistic service to people with pro­ blem s” (Bauer, 1989). ACKNOWLEDGEMENTS: I would like to thank the following persons: Zodw a Kubheka and Naumi M ashalane, for their dedication and thoroughness as research assistants; Hellen M yezw a for her friendly and collegial support; Dr. Piet Bekker for his responsiveness in advising and dou­ ble checking the statistics in this report and last but not least sincerest gratitude to G TZ - H ealth System s R esearch Eastern and Southern Africa, for bearing the full cost o f this study. REFERENCES B arrie JL, M arsh D R 1992 Q uality o f d ata in the M a n c h este r orth o p ae d ic database. B ritish M edical Journal 304 (6820); 159-62 B a u e r D 1989 F o u n d a tio n s o f P h y sic a l R e h abilitation - A M a n a g em en t A pproach. 1st Edition. C hurchill L ivingstone B uckley N A , W hyte 1M, D aw son A H , R eith D A 1999 P reform atted adm ission ch arts fo r poisoning a d m issions facilitate clinical asse ss­ m en t a nd research. A n n a ls o f E m e rg en c y M e dicine 34(4 Pt 1): 476-482. D e vitt JH , R apanosT, K urrek M , C ohen M M , S haw M 1999 T he a n esthetic record: accuracy and com pleteness. C anadian Journal o f A n a es­ thetics 46(2): 122-128 F itz m a u ric e JM 1994 H e a lth c a re d a ta standards are required fo r m edically effective use at w orkstations. International Journal o f Biom edical C om puters 34(1-4): 331-4 H aughom JL , T ransform ing U .S. H ealth Care 200 0 T he arduous ro ad to value. Top H ealth Inform atics M a n a g em en t 20(3): 1-10 H o L M , M cG hee SM , H edley AJ, L eong JC 1999 T he a p p lication o f a com p u te riz ed p roblem o rien ted m ed ic al re c o rd and it’s im pact on p a tient care. International Journal o f M edical Inform atics 55(1): 4 7-59 K ondziolka D, S c h w a rtz M L , W alters BC, M cN eil 1 1989 T he S u n n y b ro o k N eurotraum a A ssessm ent R ecord: Im proving T raum a Data C ollection. Journal o f T raum a 29(6): 730-5 L arabee JH , B o ldreghini S, E ld er-S o rrells K, T u rn e r Z M , W en d e r R G , H a rt J 2001 E valuation o f d o c u m entation before and after im plem entation o f an inform ation system in an acute care hospital. C o m p u te rize d N ursing 19(2): 56-65; quiz 66-8 M ’kum buzi VRP, E ales C J 2001. P hysio­ therapy and O c cu pational T herapy: training to g eth e r tow ard p atient care. In Press. M y e z w a H M 1997 T h e im p a c t o f C B R services in Z im babw e. A sia Pacific Journal 6(1): 7 - 9 M yezw a H M , M ’kum buzi VRP, M huri F 2001 Q uality A ssurance in a R ehabilitation Service. South A frican Journal o f P h y siotherapy 57 (1): 7-11 Perlm an BB, S chw artz A H , Paris M , T hornton JC , S m ith H, W eber R 1982 P sy ch ia tric records: variations based on disc ip lin e and patient chara cte ristics, w ith im p licatio n s for quality o f care. A m erican Journal o f Psychiatry 139(9): 1154-7 Phillips D R, H ym a B, R am esh A 1992 A c o m p a ris o n o f the use o f trad itio n a l and m odern m edicine on prim ary health care c e n ­ tres in Tam il N ada. G eo Journal 26(1): 21-30 Tarlov A R 1992 T he co m in g influence o f a social sciences perspective on m edical e d u c a ­ tion. A cadem ic M e dicine 67(11): 724-31 T h e M edical R ecords Institute. 1999. “E le c ­ tro n ic h e a lth re c o rd s tre n d s and u s a g e .” h ttp ://w w w .m e d re c in s t.c o m /s u rv e y /a n a ly s is / results.shtm ail (10 M arch 2001) S A J o u r n a l o f P h y s io t h e r a p y 2002 V o l 58 No 1 2 7 R ep ro du ce d by S ab in et G at ew ay u nd er li ce nc e gr an te d by th e P ub lis he r (d at ed 2 01 3. ) http://www.medrecinst.com/survey/analysis/