Quality accounts - methodology

What is Dr Foster Intelligence's aggregate score?

No single measure can provide the whole picture, and no single number can possibly capture the complexity of what hospitals do. Nonetheless, it is often useful to combine different measures to give each hospital a score.

For our quality accounts and scoring system we decided to concentrate on the area we think is most important - patient safety. We have used the relevant indicators from our quality accounts (ie the 13 indicators from the patient safety domain) and combined them using a statistical technique called z-scoring to give an overall measure of each hospital. No weighting was applied to the indicators.

The result is a score between 0 and 100, with 100 being the best. We have grouped all trusts into five bands, giving 1 for the poorest performers and 5 for the best.

The methodology identifies top and bottom performers by looking at their score for a certain measure of care and seeing how different this is to other hospitals. Those who are doing much better than the average get high scores and those who are not doing as well get low scores.

To put them in bandings, we have calculated the probability that hospitals are among the best or worst, based on their scores.

What methodology did we use?

Step 1 Produce individual z scores

Calculate z scores for each indicator. Different types of indicators have different methods of calculating a z score that are summarised below.

Based on the recommendation of the Healthcare Commission (2009), relative risk based indicators are transformed thus: Rates are transformed by subtracting the mean value for the rate then dividing by its standard deviation. Categorical data (e.g. yes/no question etc) are mapped onto a scale between -2.5 and +2.5. Before aggregating, all z scores are capped at ± 3 to stop a single extreme value dominating the aggregate score.

For this methodology to work, all indicators need to be in the form that high is bad. Therefore, z scores where this was not the case were multiplied by -1.

Step 2 - Produce aggregate z scores

The z-scores for each of the indicators were averaged to produce the domain z-score for patient safety. Missing values were excluded from the z-score. To ensure that the domain score behaves like proper z-scores, the score for each trust is transformed by subtracting the mean value and then dividing this by its standard deviation.

Step 3 - Produce Bayesian Ranking

Each indicator has an associated degree of uncertainty which must be taken into consideration if we are going to attempt to rank trusts. For instance, our example trust has an HSMR of 84.96073 with Poisson distributed 95% Confidence interval between 79.0 and 91.2. We used a statistical package called R to generate 50,000 credible estimates for each indicator for every trust. The graph below shows the distribution of HSMR scores for the example trust that were used in the ranking process.

In effect, Bayesian ranking repeatedly picks HSMRs from this distribution and sees how this changes the ranking. As it makes use of lots of random numbers it's often called a Monte Carlo procedure.

The key effect of the Bayesian ranking is to move trusts with extreme scores and wide confidence intervals towards the overall average.

In the School mark example below, the Bayesian ranking takes into account.

Each indicator has an associated degree of uncertainty to be incorporated into the Bayesian ranking in the following manner:

SMRs - Poisson distribution with lambda=number of deaths.
SARs - Poisson distribution with lambda=number of admissions.
Crude event rates - Depends on the rate, r;
- If 0.1 < r< 0.9 then normal approximation to the binomial distribution,
- If 0.1 >r or r>0.9 then Poisson distribution with lambda=number of events.
Crude ratio - currently the only crude ratio is the staff / bed ratio and this will be kept fixed as it is within a hospitals control.
Response to hospital guide questionnaire - kept fixed as they are within the hospitals control.

In their research, Marshall and Spiegelhalter (1998) used 10,000 iterations. Computing power has dramatically increased since 1998, so we used 50,000 iterations. To ensure this was sufficient, we ran the procedure for an additional 50,000 iterations and verified that the trusts in the top 10 and bottom 10 trusts remain unchanged for each domain indicator.

Output

The project brief was for each trust to be given a score and a banding for patient safety.

Scores were calculated by taking the trusts' median rank (which goes from 1 to 147) and rescaling it, so it lay between 0 and 100.

The bandings were designed to reflect the uncertainty over the ranking. Conventionally, median ranks are shown alongside a 75% credible region and perhaps the probability of a trust being in the top 10 and bottom 10 trusts. However, we were advised that this approach needed refining for a public guide.

Instead, trusts with a very high probability of being in the top and bottom 10 were allocated a 5 or 1 banding respectively. Trusts whose 75% credible region overlapped the mid point were allocated a banding of 3. The remainder were allocated a banding of 4 or 2 depending on whether their median rank was above or below the half way point.

Occasionally the Bayesian ranking process produced bandings for a trust was very much out of tune with its neighbours. When this occurred, we gave the trust the benefit of the doubt and promoted the trust up to a higher banding.

All my trust's data is defined as "within expected range" why have I been banded 1 in Dr Foster's analysis?

One important thing to note is that z scores only look at the point estimate of a trust's score (eg the HSMR value not the control limit banding) and without Bayesian ranking ignore the uncertainty about a score. In this respect they differ from our other reports that report indicators as low, within expected range and high. This effect is illustrated in the following example. A school may impose a pass mark of 50% across all exams. Moreover, the School may say that as marks are subjective, marks are only valid to the nearest 5%. Consider a student with the following marks:

Geography: 45% (CI 40% to 50%) - As Confidence intervals included 50% give a pass.
English: 45% (CI 40% to 50%) - As Confidence intervals included 50% give a pass.
History: 45% (CI 40% to 50%) - As Confidence intervals included 50% give a pass.
Overall mark - Pass

In this case the pass is the equivalent to "within expected range". However, their actual mark for every course was a fail so perhaps we might want to give them a fail, even though they were given a pass in every module.

We are happy to share the underlying data to interested parties. Please email qualityaccounts@drfoster.co.uk for access.

References

Bardsley, Spiegelhalter, Blunt, Chitnis, Roberts and Bharania (2009) 'Using Routine Intelligence to Target Inspection of Healthcare Providers in England', Qual. Saf. Health Care 2009; 18; 189-194.

Marshall and Spiegelhalter (1998), 'Reliability of League Tables of In Vitro Fertilisation Clinics: Retrospective Analysis of Live Birth Rates', BMJ 1998; 316; 1701-1704.

Healthcare Commission (2009), 'Following Up Mortality Outliers', www.cqc.org.uk downloaded 05/10/2009

Death in low-mortality CCS groups

Metric:

Number of deaths per 1,000 spells

Numerator

All spells with method of discharge is death (DISMETH=4)

Denominator:

Discharges, 18 years or older, spells related to pregnancy, childbirth and puerperium, or CCS groups with less than 0.5% mortality rate

Low mortality CCS groups

CCS group id	CCS group
7	Viral infection
9	Sexually transmitted infections (not HIV or hepatitis)
46	Benign neoplasm of uterus
47	Other and unspecified benign neoplasm
57	Immunity disorders
61	Sickle cell anaemia
64	Other haematologic conditions
65	Mental retardation
66	Alcohol-related mental disorders
67	Substance-related mental disorders
72	Anxiety, somatoform, dissociative, and personality disorders
73	Preadult disorders
84	Headache, including migraine
86	Cataract
87	Retinal detachments, defects, vascular occlusion, and retinopathy
88	Glaucoma
89	Blindness and vision defects
90	Inflammation, infection of eye
91	Other eye disorders
92	Otitis media and related conditions
93	Conditions associated with dizziness or vertigo
94	Other ear and sense organ disorders
119	Varicose veins of lower extremity
120	Haemorrhoids
247	Lymphadenitis
124	Acute and chronic tonsillitis
126	Other upper respiratory infections
128	Asthma
136	Disorders of teeth and jaw
137	Diseases of mouth, excluding dental
138	Oesophageal disorders
140	Gastritis and duodenitis
141	Other disorders of stomach and duodenum
142	Appendicitis and other appendiceal conditions
143	Abdominal hernia
147	Anal and rectal conditions
198	Other inflammatory condition of skin
200	Other skin disorders
202	Rheumatoid arthritis and related disease
203	Osteoarthritis
208	Acquired foot deformities
209	Other acquired deformities
211	Other connective tissue disease
212	Other bone disease and musculoskeletal deformities
225	Joint disorders and dislocations, trauma-related
160	Calculus of urinary tract
162	Other diseases of bladder and urethra
163	Genitourinary symptoms and ill-defined conditions
164	Hyperplasia of prostate
165	Inflammatory conditions of male genital organs
166	Other male genital disorders
167	Nonmalignant breast conditions
168	Inflammatory diseases of female pelvic organs
169	Endometriosis
170	Prolapse of female genital organs
171	Menstrual disorders
172	Ovarian cyst
173	Menopausal disorders
174	Female infertility
175	Other female genital disorders
177	Spontaneous abortion
178	Induced abortion
179	Postabortion complications
180	Ectopic pregnancy
181	Other complications of pregnancy
182	Haemorrhage during pregnancy, abruptio placenta, placenta previa
183	Hypertension complicating pregnancy, childbirth and the puerperium
184	Early or threatened labour
185	Prolonged pregnancy
186	Diabetes or abnormal glucose tolerance complicating pregnancy, childbirth, or the puerperium
187	Malposition, malpresentation
188	Fetopelvic disproportion, obstruction
189	Previous C-section
190	Foetal distress and abnormal forces of labour
191	Polyhydramnios and other problems of amniotic cavity
192	Umbilical cord complication
193	Trauma to perineum and vulva
194	Forceps delivery
195	Other complications of birth, puerperium affecting management of mother
196	Normal pregnancy and/or delivery
218	Liveborn
220	Intrauterine hypoxia and birth asphyxia
222	Haemolytic jaundice and perinatal jaundice
223	Birth trauma
214	Digestive congenital anomalies
215	Genitourinary congenital anomalies
217	Other congenital anomalies
102	Nonspecific chest pain
251	Abdominal pain
232	Sprains and strains
236	Open wounds of extremities
241	Poisoning by psychotropic agents
242	Poisoning by other medications and drugs
253	Allergic reactions
10	Immunizations and screening for infectious disease
176	Contraceptive and procreative management
256	Medical examination/evaluation
257	Other aftercare
258	Other screening for suspected conditions

Exclude cases

With any code for trauma, immunocompromised state, or cancer

See Appendix A: Codes for immunocompromised states

See Appendix B: Cancer codes

See Appendix C: Trauma diagnosis codes

Data Source

SUS - CDS

Time frame

April 2008 - March 2009

Basis

Acute Trust

Statistical methods used

Crude rate per1,000

Appendix A Immunocompromised states

B24	Unspecified human immunodefiency virus [HIV] disease
B59	Pneumocystosis
D70	Agranulocytosis
D7	Functional disorders of polymorphonuclear neutrophils
D72	Genetic anomalies of leukocytes
D80	Hereditary hypogammaglobulinaemia
D80	Nonfamilial hypogammaglobulinaemia
D80	Selective deficiency of immunoglobulin A [IgA]
D80	Selective deficiency of immunoglobulin G [IgG] subclasses
D80	Selective deficiency of immunoglobulin M [IgM]
D80	Immunodefiency with increased immunoglobulin M [IgM]
D80	Other immunodefiencies with predominantly antibody defects
D81	Nezelof's syndrome
D819	Combined immunodefiency, unspecified
D82	Wiskott-Aldrich syndrome
D82	Di George's syndrome
D8	Com var immunodef with predom abn B-cell numb and funct
D8	Common var immunodef predom immunoregulatory T-cell disord
D8	Other common variable immunodefiencies
D849	immunodefiency, unspecified
D89	Other specified disorders involving the immune mechanism NEC
D899	Disorder involving the immune mechanism, unspecified
E4	Kwashiorkor
E4	Nutritional marasmus
E4	Unspecified severe protein-energy malnutrition
I12	Hypertensive renal disease with renal failure
I1	Hypertensive heart and renal disease with renal failure
I1	Hyper heart and renal dis both (cong) heart and renal fail
K91	Postsurgical malabsorption, not elsewhere classified
M	59 Systemic involvement of connective tissue, unspecified
N1	Chronic renal failure (no exact ICD1 match possible)
T86	Bone-marrow transplant rejection
T86	Kidney transplant failure and rejection
T86	Heart transplant failure and rejection
T86	Liver transplant failure and rejection
T86	Failure and reject of other transplanted organs and tissues
T869	Failure and reject of unspec transplanted organ and tissue
Z45	Adjustment and management of vascular access device
Z49	Extracorporeal dialysis
Z94	Kidney transplant status
Z94	Heart transplant status
Z94	Lung transplant status
Z94	Liver transplant status
Z94	Other transplanted organ and tissue status
Z99	Dependence on renal dialysis

Appendix B Cancer codes

Whole C chapter
D00	Carcinomata-in-situ of oral cavity oesophagus and stomach
D01	Carcinomata-in-situ of other and unspecified digestive organs
D02	Carcinomata-in-situ of middle ear and respiratory system
D03	Melanoma in situ
D04	Carcinomata-in-situ of skin
D05	Carcinomata-in-situ of breast
D06	Carcinomata-in-situ of cervix uteri
D07	Carcinomata-in-situ of other and unspecified genital organs
D09	Carcinomata-in-situ of other and unspecified sites
Z85	Personal history of malignant neoplasm

Appendix C Trauma diagnosis codes

S011	Open wound of eyelid and periocular area
S02	Fracture of vault of skull
S02	Fracture of base of skull
S02	Fracture of nasal bones
S02	Fracture of orbital floor
S02	Fracture of malar and maxillary bones
S02	Fracture of mandible
S02	Multiple fractures involving skull and facial bones
S02	Fractures of other skull and facial bones
S02	Fracture of skull and facial bones, part unspecified
S03	Dislocation of jaw
S05	Ocular lacn and rupture with prolapse or loss intraoc tiss
S05	Ocular lacn without prolapse or loss of intraocular tissue
S05	Penetrating wound of orbit with or without foreign body
S05	Penetrating wound of eyeball with foreign body
S05	Avulsion of eye
S05	Other injuries of eye and orbit
S05	Injury of eye and orbit, part unspecified
S06	Concussion
S06	Diffuse brain injury
S06	Epidural haemorrhage
S06	Traumatic subdural haemorrhage
S06	Traumatic subarachnoid haemorrhage
S06	Other intracranial injuries
S07	Crushing injury of face
S09	Unspecified injury of head
S12	Fracture of first cervical vertebra
S12	Fracture of second cervical vertebra
S12	Multiple fractures of cervical spine
S12	Fracture of other parts of neck
S12	Fracture of neck, part unspecified
S13	Dislocation of cervical vertebra
S13	Multiple dislocations of neck
S14	Other and unspecified injuries of cervical spinal cord
S14	Injury of nerve root of cervical spine
S14	Injury of brachial plexus
S15	Injury of carotid artery
S15	Injury of external jugular vein
S15	Injury of internal jugular vein
S15	Injury of multiple blood vessels at neck level
S15	Injury of other blood vessels at neck level
S15	Injury of unspecified blood vessel at neck level
S17	Crushing injury of neck, part unspecified
S22	Fracture of thoracic vertebra
S22	Fracture of sternum
S22	Fracture of rib
S22	Multiple fractures of ribs
S22	Flail chest
S23	Dislocation of thoracic vertebra
S24	Other and unspecified injuries of thoracic spinal cord
S24	Injury of nerve root of thoracic spine
S25	Injury of thoracic aorta
S25	Injury of innominate or subclavian artery
S25	Injury of superior vena cava
S25	Injury of innominate or subclavian vein
S25	Injury of pulmonary blood vessels
S25	Injury of intercostal blood vessels
S25	Injury of multiple blood vessels of thorax
S25	Injury of other blood vessels of thorax
S25	Injury of unspecified blood vessel of thorax
S26	Other injuries of heart
S26	Injury of heart, unspecified
S27	Traumatic pneumothorax
S27	Traumatic haemothorax
S27	Traumatic haemopneumothorax
S27	Other injuries of lung
S27	Injury of bronchus
S27	Injury of other specified intrathoracic organs
S27	Injury of unspecified intrathoracic organ
S32	Fracture of lumbar vertebra
S32	Fracture of sacrum
S32	Fracture of ilium
S32	Fracture of acetabulum
S32	Fracture of pubis
S32	Fracture of oth and unspec parts of lumbar spine and pelvis
S33	Dislocation of lumbar vertebra
S33	Dislocation of sacroiliac and sacrococcygeal joint
S34	Other injury of lumbar spinal cord
S34	Injury of nerve root of lumbar and sacral spine
S34	Injury of cauda equina
S34	Injury of lumbosacral plexus
S35	Injury of abdominal aorta
S35	Injury of inferior vena cava
S35	Injury of coeliac or mesenteric artery
S35	Injury of portal or splenic vein
S35	Injury of renal blood vessels
S35	Injury of iliac blood vessels
S35	Inj multi blood vessels abdomen lower back and pelvis level
S35	Inj oth blood vessels abdomen lower back and pelvis level
S35	Inj of unspec blood vessel abdom low back and pelvis level
S36	Injury of spleen
S36	Injury of liver or gallbladder
S36	Injury of pancreas
S36	Injury of stomach
S36	Injury of small intestine
S36	Injury of colon
S36	Injury of rectum
S36	Injury of other intra-abdominal organs
S36	Injury of unspecified intra-abdominal organ
S37	Injury of kidney
S37	Injury of ureter
S37	Injury of bladder
S37	Injury of uterus
S37	Injury of other pelvic organs
S37	Injury of unspecified pelvic organ
S38	Crushing injury of external genital organs
S38	Crush inj oth and unspec part of abdo lower back and pelvis
S42	Fracture of clavicle
S42	Fracture of scapula
S42	Fracture of upper end of humerus
S42	Fracture of shaft of humerus
S42	Fracture of lower end of humerus
S43	Dislocation of shoulder joint
S43	Dislocation of acromioclavicular joint
S43	Dislocation of sternoclavicular joint
S43	Dislocation of other and unspec parts of shoulder girdle
S45	Injury of brachial artery
S45	Injury of axillary or brachial vein
S4	Crushing injury of shoulder and upper arm
S52	Fracture of upper end of ulna
S52	Fracture of upper end of radius
S52	Fracture of shaft of ulna
S52	Fracture of shaft of radius
S52	Fracture of shafts of both ulna and radius
S52	Fracture of lower end of radius
S52	Fracture of lower end of both ulna and radius
S52	Fracture of other parts of forearm
S52	Fracture of forearm, part unspecified
S53	Dislocation of elbow, unspecified
S55	Injury of vein at forearm level
S57	Crushing injury of elbow
S57	Crushing injury of forearm, part unspecified
S58	Traumatic amputation at elbow level
S58	Traumatic amputation at level between elbow and wrist
S62	Fracture of navicular [scaphoid] bone of hand
S62	Fracture of other carpal bone(s)
S62	Fracture of first metacarpal bone
S62	Fracture of other metacarpal bone
S62	Multiple fractures of metacarpal bones
S62	Fracture of other and unspecified parts of wrist and hand
S63	Dislocation of wrist
S63	Dislocation of finger
S65	Injury of superficial palmar arch
S65	Injury of blood vessel(s) of other finger
S67	Crushing injury of thumb and other finger(s)
S67	Crush injury other and unspecified parts of wrist and hand
S71	Open wound of hip
S72	Fracture of neck of femur
S72	Pertrochanteric fracture
S72	Subtrochanteric fracture
S72	Fracture of shaft of femur
S72	Fracture of lower end of femur
S72	Fracture of femur, part unspecified
S73	Dislocation of hip
S75	Injury of femoral artery
S75	Injury of femoral vein at hip and thigh level
S75	Injury of greater saphenous vein at hip and thigh level
S77	Crushing injury of hip with thigh
S78	Traumatic amputation of hip and thigh, level unspecified
S81	Open wound of knee
S82	Fracture of patella
S82	Fracture of upper end of tibia
S82	Fracture of shaft of tibia
S82	Fracture of fibula alone
S82	Fracture of medial malleolus
S82	Fracture of lateral malleolus
S82	Fractures of other parts of lower leg
S83	Dislocation of patella
S83	Dislocation of knee
S83	Tear of meniscus, current
S83	Tear of articular cartilage of knee, current
S85	Injury of popliteal artery
S85	Injury of (anterior)(posterior) tibial artery
S85	Injury of popliteal vein
S85	Injury of other blood vessels at lower leg level
S85	Injury of unspecified blood vessel at lower leg level
S87	Crushing injury of knee
S87	Crushing injury of other and unspecified parts of lower leg
S88	Traumatic amputation at knee level
S88	Traumatic amputation at level between knee and ankle
S88	Traumatic amputation of lower leg, level unspecified
S91	Open wound of other parts of foot
S92	Fracture of calcaneus
S92	Fracture of talus
S92	Fracture of other tarsal bone(s)
S92	Fracture of metatarsal bone
S92	Fracture of foot, unspecified
S93	Dislocation of ankle joint
S93	Dislocation of toe(s)
S93	Dislocation of other and unspecified parts of foot
S95	Injury of plantar artery of foot
S95	Injury of other blood vessels at ankle and foot level
S95	Injury of unspecified blood vessel at ankle and foot level
S97	Crushing injury of ankle
S97	Crushing injury of toe(s)
S97	Crushing injury of other parts of ankle and foot
S98	Traumatic amputation of foot at ankle level
T01	Open wounds involving multiple regions of upper limb(s)
T02	Fractures involving thorax with lower back and pelvis
T02	Fractures involving multiple regions of both upper limbs
T02	Fractures involving multiple regions of both lower limbs
T03	Multiple dislocations, sprains and strains, unspecified
T04	Crush inj involv thorax with abdomen lwr back & pelvis
T04	Crushing injuries involving multiple region of upper limb(s)
T04	Crushing injuries involving multiple region of lower limb(s)
T04	Multiple crushing injuries, unspecified
T05	Traumatic amputation of both arms [any level]
T05	Traumatic amputation of both feet
T05	Traumatic amputation of both legs [any level]
T06	Inj nerves and spinal cord involving oth multi body regions
T08	Fracture of spine, level unspecified
T08	Fracture of spine, level unspecified
T09	Dislocation sprain & strain unspec joint & ligament trunk
T09	Injury of spinal cord, level unspecified
T09	inj unspec nerve spinal nerve root & plexus trunk
T10	Fracture of upper limb, level unspecified
T10	Fracture of upper limb, level unspecified
T11	Injury of unspec blood vessel of upper limb level unspec
T11	Traumatic amputation of upper limb, level unspecified
T12	Fracture of lower limb, level unspecified
T12	Fracture of lower limb, level unspecified
T13	Open wound of lower limb, level unspecified
T14	Fracture of unspecified body region
T14	Dislocation, sprain and strain of unspecified body region
T14	Injury of blood vessel(s) of unspecified body region
T14	Crush injury and traumatic amputation of unspec body region
T14	Other injuries of unspecified body region
T20	Burn of unspecified degree of head and neck
T20	Burn of first degree of head and neck
T20	Burn of second degree of head and neck
T20	Burn of third degree of head and neck
T21	Burn of unspecified degree of trunk
T21	Burn of first degree of trunk
T21	Burn of second degree of trunk
T21	Burn of third degree of trunk
T22	Burn unspec degree should and upp limb excpt wrist and hand
T22	Burn first deg of shoulder and up limb excpt wrist and hand
T22	Burn sec deg of shoulder and upr limb exc wrist/hand
T22	Burn third deg shoulder and upper limb excpt wrist and hand
T23	Burn of unspecified degree of wrist and hand
T26	Burn of eyelid and periocular area
T26	Burn of cornea and conjunctival sac
T26	Burn with resulting rupture and destruction of eyeball
T26	Burn of eye and adnexa, part unspecified
T79	Air embolism (traumatic)
T79	Fat embolism (traumatic)
T79	Traumatic secondary and recurrent haemorrhage
T79	Post-traumatic wound infection, not elsewhere classified
T79	Traumatic shock
T79	Traumatic anuria
T79	Traumatic ischaemia of muscle
T79	Traumatic subcutaneous emphysema
T79	Other early complications of trauma

Operated within 2 days

Metric: Percentage of operations performed within 2 days of admission for diagnosis of fractured neck of femur
Numerator: Spells where opdate minus admission date less than 2
Denominator: All spells with primary diagnosis of fractured neck of femur where opdate is not null
ICD10 codes: Fracture of neck of femur (hip)- S720-S722
Exclude cases: Opdate is null
Data Source: SUS - CDS
Time frame: April 2008 - March 2009
Basis: Acute Trust
Statistical methods used: Crude rate

Hip and Knee revision rates with manipulations

Metric

The proportion of joint replacements with a revision procedure within 365 days of the initial (index) procedure, over the total number of joint replacements carried out at the trust over a three year period.

Three years of index procedures are combined to provide sufficient numbers at trust level. A further year of data is needed to allow a year's follow-up for every index procedure. Index operations in 2005/6 to 2007/8 give rise, potentially, to revisions between 2005/6 and 2008/9.

Note that only one revision within 365 days per patient is counted (some people can have several), and revisions are matched to side of index procedure (right or left).

Construction

Index procedure codes for primary total hip replacement (THR)

Codes: oper1 in ('W371','W381','W391') or (oper1='W581' and (oper2='Z843' or oper3='Z843' or oper4='Z843'))

Description: These cover primary total replacement including resurfacing procedures. For the latter, the Z code is required in any secondary op field (oper2-oper12 - only 2 to 4 shown above for brevity) to specify hip.

Revision procedure codes for THR

Codes: oper1 in ('W373','W383','W393', 'W372','W374', 'W382','W384','W392','W394') or (oper1='W582' and (oper2='Z843' or oper3='Z843' or oper4='Z843')) or oper2 in ('W580', 'W370', 'W380', 'W390') or oper3 in ('W580', 'W370', 'W380', 'W390') or oper4 in ('W580', 'W370', 'W380', 'W390');

Description: These cover revisions (primary op field), conversions (in any secondary op field) and "attention to joint" (primary op field).

Index procedure codes for primary total knee replacement (TKR)

Codes: oper1 in ('W401','W411','W421') or (oper1='W581' and (oper2='Z846' or oper3='Z846' or oper4='Z846'))

Description: These cover primary total replacement including resurfacing procedures. For the latter, the Z code is required in any secondary op field to specify knee.

Revision procedure codes for TKR

Codes: oper1 in ('W403','W413','W423', 'W402','W404', 'W412','W414', 'W422') or ('W424' and 'Y032') or (oper1 in ('W580','W582') and (oper2='Z846' or oper3='Z846' or oper4='Z846')) or oper2 in ('W400', 'W410', 'W420') or oper3 in ('W400', 'W410', 'W420') or oper4 in ('W400', 'W410', 'W420')

Description: These cover revisions (primary op field), conversions (in any secondary op field) and "attention to joint" (primary op field).

Data Source

SUS - CDS

Time frame

Index procedure- April 2005 - March 2008

Revisions- April 2005 - March 2009

Basis

Acute Trust

Statistical methods used

Crude rate

Banding

Upper and lower control limits are calculated at 99.8% level of significance and these are used to band the crude rate.

There are three reported bands:

if the crude rate of revisions is greater than the upper control limit then the assigned band is red, "more than expected"
if the crude rate of revisions is less than the lower control limit then the assigned band is red, "less than expected"
otherwise, the assigned band is blue, "as expected"

Median outpatient waiting time (days)

Metric

Median outpatient waiting time (days)

Calculation

Median waiting time ('Waiting' gives the period in days between the date of the appointment date and either the referral request received date (reqdate) or the DNA (did not attend) date, if given.) for each Trust.

Criteria

Acute specialties only (see appendix)
Patients referred by a General Medical Practitioner, Dentist or Dental service
Patients with a null or zero waiting time are excluded
First attendances

Data Source

SUS - CDS

Time frame

April 2008 - March 2009

Basis

Acute Trust

Statistical methods used

Z score

Appendix

Group id	Diagnosis group
100	General Surgery
10	Urology
10	Transplantation Surgery
10	Breast Surgery
10	Colorectal Surgery
10	Hepatobiliary & Pancreatic Surgery
10	Upper Gastrointestinal Surgery
10	Vascular Surgery
110	Trauma & Orthopaedics
120	ENT
130	Ophthalmology
140	Oral Surgery
14	Restorative Dentistry
14	Orthodontics
14	Maxillo-Facial Surgery
14	Oral & Maxillo Facial Surgery
14	Endodontics
14	Periodontics
14	Prosthodontics
14	Surgical Dentistry
150	Neurosurgery
160	Plastic Surgery
16	Burns Care
170	Cardiothoracic Surgery
17	Cardiac Surgery
17	Thoracic Surgery
17	Cardiothoracic Transplantation
180	Accident & Emergency
190	Anaesthetics
19	Pain Management
19	Critical Care Medicine
14	Paediatric Dentistry
17	Paediatric Surgery
300	General Medicine
30	Gastroenterology
30	Endocrinology
30	Clinical Haematology
30	Clinical Physiology
30	Clinical Pharmacology
30	Hepatology
30	Diabetic Medicine
30	Blood And Marrow Transplantation
30	Haemophilia
310	Audiological Medicine
31	Clinical Genetics
31	Clinical Cytogenetics And Molecular Genetics
31	Clinical Immunology And Allergy
31	Rehabilitation
31	Palliative Medicine
31	Clinical Immunology
31	Allergy
31	Intermediate Care
31	Respite Care
320	Cardiology
32	Clinical Microbiology
330	Dermatology
340	Respiratory Medicine
34	Respiratory Physiology
350	Infectious Diseases
35	Tropical Medicine
360	Genito-Urinary Medicine
36	Nephrology
370	Medical Oncology
37	Nuclear Medicine
400	Neurology
40	Clinical Neuro-Physiology
410	Rheumatology
450	Dental Medicine Specialties
460	Medical Ophthalmology
32	Paediatric Cardiology
420	Paediatrics
42	Paediatric Neurology
620	GP Non-maternity
800	Clinical Oncology (previously Radiotherapy)
810	Radiology
81	Interventional Radiology
820	General Pathology
82	Blood Transfusion
82	Chemical Pathology
82	Haematology
82	Histopathology
830	Immunopathology
83	Medical Microbiology
83	Neuropathology
900	Community Medicine
90	Occupational Medicine

Number of daycase overstays

Metric

Number of daycase overstays

Numerator

Spells where management intent =2 (day case) and length of stay is greater than zero.

Denominator

All spells where management intent =2 (day case)

Exclude cases

Elective only- admission method:

Data Source

SUS - CDS

Time frame

April 2008 - March 2009

Basis

Acute Trust

Statistical methods used

Crude rate

Hospital Standardised Mortality Rates- AMI

Metric

The ratio of the observed number of in-hospital deaths to the expected number of deaths, multiplied by 100.

Numerator

All spells with method of discharge as death (DISMETH=4), defined by a specific diagnosis code for the primary diagnosis of the spell.

Diagnosis group

Acute myocardial infarction- ICD10 codes: I21,I22

Denominator

Expected number of in-hospitals deaths derived from logistic regression.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Exclude cases

Daycases (where classpat = 2 in the first episode)

Data Source / Time frame

SUS - April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

The ratio is calculated by dividing the actual number of deaths by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has an HSMR of 100, that means that the number of patients who died is exactly as it would be expected taking into account the standardisation factors. An HSMR above 100 means more patients died than would be expected; one below 100 means that fewer than expected died.

Control limits

Control limits tell us the range of values which are consistent with random or chance variation. Data points falling within the control limits are consistent with random or chance variation and are said to display 'common-cause variation'; for data points falling outside the control limits, chance is an unlikely explanation and hence they are said to display 'special-cause variation' - that is, where the trust's rate diverges significantly from the national rate.

Hospital Standardised Mortality Rates- FNOF

Metric

The ratio of the observed number of in-hospital deaths to the expected number of deaths, multiplied by 100.

Numerator

All spells with method of discharge as death (DISMETH=4), defined by a specific diagnosis code for the primary diagnosis of the spell.

Diagnosis group

Fracture of neck of femur (hip)- ICD10 codes: S720-S722

Denominator

Expected number of in-hospitals deaths derived from logistic regression.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Exclude cases

Daycases (where classpat = 2 in the first episode)

Data Source / Time frame

SUS - April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Hospital Standardised Mortality Rates- Stroke

Metric

The ratio of the observed number of in-hospital deaths to the expected number of deaths, multiplied by 100.

Numerator

All spells with method of discharge as death (DISMETH=4), defined by a specific diagnosis code for the primary diagnosis of the spell.

Diagnosis group

Acute cerebrovascular disease- ICD10 codes: G46,I60-I64,I66

Denominator

Expected number of in-hospitals deaths derived from logistic regression.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Exclude cases

Daycases (where classpat = 2 in the first episode)

Data Source / Time frame

SUS - April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Standardised Readmission Ratio- Cholecystectomy

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Procedure Groups-

Laparascopic cholecystectomy- OPCS codes: J183 (+ Y508 or Y75)

Other excision of gall bladder- OPCS codes: J18 (excluding J183 with Y508 or Y75)

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

The ratio is calculated by dividing the actual number of readmissions by the expected number and multiplying the figure by 100. It is expressed as a relative risk, where a risk rating of 100 represents the national average. If the trust has a SRR of 100, that means that the number of patients who were readmitted is exactly as would be expected taking into account the standardisation factors. A SRR above 100 means more patients were readmitted than would be expected; one below 100 means that fewer than expected were readmitted.

Control limits

Standardised Readmission Ratio- Fracture of neck of femur

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Diagnosis Group-

Fracture of neck of femur (hip)- ICD10 codes: S720-S722

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Standardised Readmission Ratio- Hip replacement

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Procedure Group-

Hip replacement- OPCS codes: W37-W39,W93-W95

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Standardised Readmission Ratio- Hysterectomy

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Procedure Group-

Abdominal excision of uterus- OPCS code: Q07

Vaginal excision of uterus- OPCS code: Q08

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Standardised Readmission Ratio- Urinary tract infections

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Diagnosis Group-

Urinary tract infections- IC10 codes: N10,N11,N151,N158,N159,N16,N291,N30,N330,N34,N351,N37,N390,P393

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Standardised Readmission Ratio- Coronary atherosclerosis and other heart disease

Metric

The ratio of the observed number of readmissions to the expected number of readmissions, multiplied by 100.

Numerator

All spells with an emergency readmission within 28 days of discharge.

Readmitting episode- Emergency admissions:

21, Emergency - via A&E
22, Emergency - via GP
23, Emergency - via Bed Bureau
24, Emergency - via Out-patient clinic
28, Emergency - via other means

Readmission date within 28 days of discharging spell

Diagnosis Group-

Coronary atherosclerosis and other heart disease- ICD10 codes: I20,I24,I251,I252,I255-I259

Denominator

Expected number of readmission derived from logistic regression, adjusting for factors to indirectly standardise for differences in case-mix.

Adjustments are made for:

Sex
Age on admission (in five year bands up to 90+)
Admission method (non-elective or elective)
Socio-economic deprivation quintile of the area of residence of the patient (based on the Carstairs Index)
Primary diagnosis (based on the Clinical Classification System - CCS group)
Co-morbidities (based on Charlson Score)
Number of previous emergency admissions
Year of discharge (financial year)
Palliative care (whether the patient is being treated in specialty of palliative care)

Data Source / Time frame

April 2008- March 2009

Basis

Acute Trust

Statistical methods used

Logistic regression

Control limits

Surgical Techniques- Gall Bladder

Metric

Standardised number of patients receiving open surgery for removal of gall bladder.

Methodology

Patients that received either a Laparoscopic cholecystectomy or other excision of gall bladder were obtained over the period 2004-2008 for analysis. The following codes were used:

Laparoscopic cholecystectomy OPCS codes - J183 (+ Y508 or Y75)

Other excision of gall bladder OPCS codes - J18 (excluding J183 with Y508 or Y75)

Logistic regression was applied to the data looking at the following factors:

Age
Sex
Deprivation (not significant in the final model)
Diagnosis group
Charlson score
Admission type

Admissions that did not have a valid age, sex, admission type or deprivation score were excluded from the analysis.

The diagnosis groups used in the analysis were:

Cancer of other GI organs, peritoneum - ICD10 codes: C17,C23,C24,C26,C451,C48,D014,D017,D019

Other and unspecified benign neoplasm- ICD10 codes D10-D24,D27-D36

Other nutritional, endocrine, and metabolic disorders- ICD10 codes D76,E65-E68-E74,E76,E77,E79,E80,E83,E85,E88,E898,E899,E90,R62,R631,R633-R635,R638,R760,R768,R769

Biliary tract disease- ICD10 codes K80-K83,K870,R932

Pancreatic disorders (not diabetes)- ICD10 codes B252,B263,K85,K86,K871,K903

Abdominal pain- ICD10 code R10

All other activity that did not fall into the diagnosis groups above were grouped together in a group called 'Other'.

The age groups used in the analysis were:

15-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
Over 85's

Data Source

SUS - CDS

Time frame

April 2008 - March 2009

Basis

Acute trust

Statistical methods used

Logistic regression was applied to the data. Stepwise regression was carried out and no variables were removed from the model.

The data was standardised for

Age
Sex
Deprivation (not significant in the final model)
Diagnosis group
Charlson score
Admission type

Surgical Techniques- Hysterectomy

Metric

Standardised number of patients receiving an abdominal hysterectomy.

Methodology

Patients that received either an elective vaginal hysterectomy or an elective abdominal hysterectomy were obtained for analysis. The following codes were used:

Abdominal hysterectomy - OPSC code Q07

Vaginal hysterectomy - OPCS code Q08

Logistic regression was applied to the data looking at the following factors:

Age
Deprivation (not significant in the final model)
Diagnosis group
Charlson score

Admissions that did not have a valid age or deprivation score were excluded from the analysis.

The diagnosis groups used in the analysis were:

Endometriosis - ICD10 code N80

Uterine Fibroids - ICD10 code D25

Malignancies- ICD10 codes C51-C57, D06, D07

Menstrual Irregularities- ICD10 codes N92-N94

The age groups used in the analysis were:

18-44
45-64
Over 65

Data Source

SUS - CDS

Time frame

April 2008 - March 2009

Basis

Acute trust

Statistical methods used

Logistic regression was applied to the data. Stepwise regression was carried out and deprivation was not found to be significant and removed from the model.

The data was standardised for:

age
diagnosis group
charlson score