Birth certificate uk copy

We have assessed the usefulness of a regional hip register in the assessment of the outcome of primary total hip replacement (THR). Over 97% of THRs performed in the Trent region in 1990 were captured onto the register and the inaccuracies recorded were less than 1.8%. In an independent assessment of 2111 patients five years after THR, 85.9% of those available for assessment responded, and 66.8 % agreed to an assessment.

The cost of this independent assessment at five years, utilising a regional hip register, was approximately L50 per implant. This is a reasonable outlay to identify problems early. Some form of registration and outcome assessment should be performed on a national level.

J Bone Joint Surg [Br] 2000;82-B:944-7.

Received 17 November 1999; Accepted after revision 6 March 2000

Recent reports of the early failure of the 3M Capital hip prosthesis1 have led to a heightened awareness by the media, the general public and government of the potential problems in joint replacement. Prostheses can be introduced widely before stringent follow-up can prove their outcome. When the early failures of the Capital hip were identified, there was little information as to which patients had received the implant, or their whereabouts. As a result attention has been focused on improving surveillance of hip replacement and its clinical outcome. There have been a number of calls in the past for the establishment of a national hip register,2-4 which have been emphasised in the light of the Capital hip failures.5

National hip registers are established in Norway and Sweden and have been effective in providing quality assurance to hip replacements.6-8 They have identified prostheses performing poorly9 and problems with certain materials such as cement,10 by registering revisions of the primary replacement. In the UK, experience is restricted to a small number of domestic registers or audit projects concerned with hip replacement. The largest and longest running of these is the Trent Regional Arthroplasty Study.

Practical problems affecting the reliability of such largescale registers are compliance of those involved, i.e., surgeons, hospitals and patients,4 the validity and completeness of the data and cost.11 Our experience with the Register, and in particular with its capacity to assess independently the five-year outcome of hip replacement in Trent, has enabled us to assess each of these factors in detail.

In this study we have assessed the accuracy and validity of the data held on the register and the practicalities of its use in a large and diverse population.

Methods

The Trent Regional Arthroplasty Study (TRAS). The TRAS is based at the Glenfield General NHS Trust, Leicester and is the only register of its kind in England. Since the beginning of 1990 and with the agreement of all the consultant orthopaedic surgeons in the region, all primary total hip and knee replacements performed throughout Trent were recorded prospectively. The Trent region has a population of 4.7 millon (1990 census), served by 22 NHS hospitals. Patients are entered onto the register by the operating surgeon who completes a standard single-page form at the time of the operation which records demographic, medical and operative details for each patient and implant. Completed forms are returned in batches to the study centre where the data are entered onto a computerised database (Epi-info software, version 6.01) by two clerks; each entry receives a unique identifying number. Each entry form is retained and filed in order to serve as a backup copy of the register.

A detailed five-year clinical assessment of the 1990 cohort of primary total hip replacements (THRs) registered with the TRAS during 1990, was carried out in 1995. During that year, 2111 THRs were performed on 2036 patients, under the care of 66 consultants, in 19 NHS and seven private hospitals. An attempt was made to assess all living patients from this group at clinics set up specifically in the NHS hospitals in Trent and with financial help from the Department of Health. Particular attention was given to verification of data, contact and tracing of patients, and hospital contact and clinic arrangements. Each of these areas will be discussed in detail.

Data verification. The TRAS employs a full-time peripatetic clerk who visits every hospital in the region to scrutinise records and to ensure that all primary hip and knee replacements performed are included on the register. Every patient is identified individually and any not already included on the register are assigned a unique identification number and a primary entry form. A copy of this form is then sent to the consultant orthopaedic surgeon who performed the operation in order that any missing data can be added. A similar system is in place for the completion of primary forms which are returned to the study centre with incomplete data. As part of the five-year follow-up study further verification and validation of data were performed. This involved checking manually through the computerised database for inaccurate, duplicate and incomplete entries. Any such records were verified against the original entry form.

The data already held on the database for the 1990 cohort have also been validated and verified against the patients' medical records. Four particular categories of data were chosen since they are vital if a register is to be useful, and they are easily available from the medical notes (Table I). Once identified, inaccurate entries were checked against the original entry form to establish whether the errors were due to incorrect completion of the form or to typographical mistakes.

Contact and tracing of patients. At the start of the fiveyear review in 1990, 130 patients (135 THRs) were known to have died. All patients thought to be alive were sent a standard letter to the address held on the register. Those who did not reply were sent a reminder one month after the initial mailing. Patients who were thought to have moved were written to at their new address if known or, failing that, their correct address was sought via the Patient Administration System (PAS) of the hospital at which the operation had been performed or the relevant Family Health Service Authority (FHSA). Ultimately, non-responding patients were registered with the Office of National Statistics (ONS) which provided a tracing service. This established whether the patients were dead or alive, and, if still alive, their whereabouts.

Hospital contact and clinic set-up. We intended to see each patient at the hospital at which the primary operation was performed. In addition to the clinic space, nursing, radiological and administrative support was required for each clinic in the 19 NHS hospitals. The hospitals in the private sector were excluded because of logistical difficulties. The arrangements were made through the orthopaedic business manager of each hospital. An initial estimate of the number of clinics required, based on the ability to assess ten patients in a clinic of half a day was calculated for each hospital. All hospitals raised the issue of financial charges to provide their services and each was requested to provide a written estimate.

Results

Data verification: capture of the THRs onto the register. At the start of the five-year outcome study, the TRAS database contained 2104 THRs for 1990. Confirmatory checking revealed eight duplicate entries (0.37%). Subsequent verification against the medical records of each THR revealed that there were a further 40 entries on the database (1.86%) which, for various reasons, should not have been included in the 1990 cohort of primary cases (Table II). Therefore, after these checks, 48 (2.2%) THRs originally included in the database were excluded, leaving 2104. More profound scrutiny of patients' records, however, revealed a further seven primary THRs (0.33%) performed during 1990, which had not been previously registered. The total verified number of THRs performed in 1990 was therefore 2111.

Data verification: accuracy of the data. It was not possible to obtain the case notes of the 82 operations performed in the private hospitals, leaving 2029 THRs with records potentially available for scrutiny. Of these, 1954 (96.3%) were actually reviewed. The percentage of incorrect or incomplete data ranged from 0.7% (date of birth) to 1.8% (prosthesis type) (Table I). Scrutiny of the original forms of each of these inaccurate entries sent to the register revealed that the form did contain the correct data, for between 5% (operative side) and 47% (data of operation), indicating, as expected, that errors occur both when the form is initially completed and during entry into the database.