| Objective: To improve and assess the patient satisfaction questionnaires ingeneral hospital as an objective instrument of measurement for evaluating the qualityof medical care and meeting the needs of modernized administration in the generalhospital.Subjects: The 2362 outpatients and 2039 inpatients from 18 general hospitalswere surveyed.Methods: The survey was carried out on the spot without any restrictions, suchas age and diseases.Reliability and validity were evaluated in this study. Cronbach's alpha wasapplied as a measure of internal consistency. The correlation between item of 'generalsatisfaction' and the total score can be verified reliability further.Factor analysis was used to evaluate construct validity with factor loading andthe contribution rate of cumulative variance. Appropriateness of the factor model wastested with Kaiser-Meyer-Olkin measure; principal components and maximumlikelihood were adopted to extracted factors followed by promax and direct obliminrotation. Only the factors with an eigenvalue equal to or greater than 1 were rotated,items with loading of 0.3 or greater on factors were retained.The correlation of the item with its own dimension as well as with otherdimensions was assessed for convergent validity and discriminant validity. With the difference from the traditional method the score of a dimension was calculated bythe sum of items belonging to this dimension excluding the item what wasconcerned with the correlation.Demographic variables including age, sex, identity and cost were also appliedfor examining construct validity. The comparison between different hospitals wasused as the auxiliary examining to the questionnaires.Data inputting was processed by Epidata3.0 database.The statistic analysis was conducted by SPSS 13.0 software.Result:OPSQ: The Outpatient Satisfaction Questionnaire (OPSQ) was consisted of 25items and 7 dimensions which were the medical cost, doctor and nursing services, waiting time, check and the auxiliary department services, treatment outcome, environment and facilities, informed content. Seven dimensions of primaryquestionnaire unchanged, but two items of 'treatment outcome' were increased tothree, and four items of 'check and the auxiliary department services' were reduced tothree, item Bll was being adjusted.There's a good retrieval rate of 100%, the validity rate was 97.75%, and totalitems response rate was 98.83%. Only B6 and B21 got low response rate (97.29% and97.50%).Cronbach's alpha coefficient for total questionnaire was fairly high (0.872). Alldimensions except the 'informed content' had a Cronbach's alpha coefficient was0.452 were also internally consistent with Cronbach's alpha coefficient greater than0.6 (from 0.607 to 0.689). The detail was the medical cost (0.689), doctor and nursingservices (0.661), waiting time (0.629), check and the auxiliary department services(0.622), treatment outcome (0.607), environment and facilities (0.633). There was asignificant positive correlation between item of 'general satisfaction' and total score(Spearman correlation coefficient is 0.620, P＜0.0001).The data suitable for factor analysis with KMO measure was 0.892. Principalcomponents and Direct Oblimin rotation methods were adopted to extracted factors, the contribution rate of cumulate variance of anterior seven factors was 57.25%. Theresult of factor analysis was basically consistent the hypothesis with the exception ofseven items (B10, B12, B13, B14, B21, B23and B24).The Spearman correlation coefficient between one dimension and the otherswas less than the Cronbach's alpha coefficient of this dimension; The Spearmancorrelation coefficient between dimension's score and total score was less than theCronbach's alpha coefficient of all items as well as was greater than the correlationcoefficient between dimensions, the internal relativity was stronger than the exterior.With traditional analyze method, the relativity between item and its own dimensionwas greater than the relativity between this item and the other dimensions. With theanalyze method in addition to this item, the relativity of most items and their owndimension was also greater than the relativity between those items and the otherdimensions except eight items (B1, B9, B10, B13, B15, B18, B21, B24).There's a significant difference; of outpatient satisfaction among hospitals(F=9.736, P=0.000). In the results of medical service satisfaction, the mostdissatisfied dimension is 'medical cost', followed by the 'waiting time'.Demographic analysis show that: patient satisfaction in marital status, familyincome, whether the first time have no significant difference(P＞0.05), on the otherhand, there were significant difference in gender(t=-2.985, P=0.003), age(F=2.486, P=0.030), occupation (F=2.544, P=0.009), education (F=5.464, P=0.000), residence(t=-2.555, P=0.011) and paid category (F=6.234, P=0.000).IPSQ: The Inpatient Satisfaction. Questionnaire (IPSQ) was consisted of 41items and 9 dimensions which were admission process, doctor services, auxiliarydepartment services, treatment outcome, medical cost, food supply, nursing care, environment and facilities, know choice. Compared to the primary questionnairewhich consisted of 39 items and 8 dimensions, there were big changes: a newdimension 'know choice' was added which include four items, The items of'treatment outcome' were increased from 2 to 4, the items of 'doctor services' werereduced from 7 to 6, the items of 'auxiliary department services' were reduced from 6 to 5, the items of 'food supply' were reduced from 5 to 3.There's a good retrieval rate of 100%, the validity rate was 97%, and total itemsresponse rate was 96.8%. There was a low response rate of individual item were B12(82.8%), B17 (83%) and B34 (82.7%), all those items were belong to the 'foodsupply' dimension.Cronbach's alpha coefficient of total questionnaire was fairly high (0.912). Alldimensions except 2 were also internally consistent with Cronbach's alpha coefficientfrom 0.669 to 0.801. That was doctor services (0.733), auxiliary department services(0.756), treatment outcome (0.727), food supply (0.750), nursing care (0.793), environment and facilities (0.801), know choice (0.669). The dimension of'admission process' and 'medical cost' had no high enough internal consistency withCronbach's alpha coefficient (0.419 and 0.378 respectively). There was a significantpositive correlation between item of 'general satisfaction' and total score (Spearmancorrelation coefficient is 0.633, P＜0.0001).The data suitable for factor analysis with KMO measure was 0.940. Maximumlikelihood and Promax rotation methods were adopted to extracted factors, thecontribution rate of cumulate variance of anterior nine factors was 54.002%. Theresult of factor analysis was basically consistent the hypothesis with the exception ofsix items (B8, B16, B20, B22, B37 and B40).Except individual data, the Spearman correlation coefficient between onedimension and the others was less than the Cronbach's alpha coefficient of thisdimension; The Spearman correlation coefficient between dimension's score and totalscore was less than the Cronbach's alpha coefficient of all items as well as wasgreater than the correlation coefficient between dimensions, the internal relativity wasstronger than the exterior. With traditional analyze method, the relativity betweenitem and its own dimension was greater than that it with other dimensions, With theanalyze method without this item, the relativity of most items and their owndimension was also greater than the relativity between those item and the otherdimensions except eight items (B3, B4, B8, B16, B20, B22, B37, B38, B40). There's a significant difference of inpatient satisfaction among hospitals(F=11.165, P=0.000). In the results of medical service satisfaction, the mostdissatisfied dimension is 'medical cost'. This is the same with outpatient.Demographic analysis show that: patient satisfaction in family income, whether the first time have no significant difference(P＞0.05), on the other hand, there were significant difference in gender(t=-3.969, P=0.000), age(F=11.672, P=0.000), marital status(F=2.686, P=0.045), occupation(F=3.936, P=0.000), education (F=8.104, P=0.000), reside, nce (t=-3.230, P=0.001) and paid category(F=3.327, P=0.000).Discussion:OPSQ: According to the evaluation result by using analyze method in additionwith this item and the factor analyze result, the new calculated Cronbach's alphacoefficient was 0.668 if combined 4 items (B1, B9, B12, B14) into one dimension. Ifcombined B24 and the 'doctor and nursing services' items B3, B25 into onedimension, the new calculated Cronbach's alpha coefficient was 0.593. If move itemB13 into 'treatment outcome' dimension, the new calculated Cronbach's alphacoefficient was 0.686. If delete B13 item from dimension 'medical cost', the newcalculated Cronbach's alpha coefficient was 0.668. Those adjustments can get betterresult.IPSQ: According to the evaluation result by using analyze method in additionwith this item and the factor analyze result, if move item B40 from dimension 'doctorservices' into 'know choice', the new calculated Cronbach's alpha coefficient on'doctor services' was up from 0.733 to 0.768, and the new calculated Cronbach'salpha coefficient on 'know choice' was up from 0.669 to 0.698. If move item B20from dimension 'admission process' into 'auxiliary department services', the newcalculated Cronbach's alpha coefficient: was a little up from 0.756 to 0.793. If moveitem B22 from dimension 'admission process' into 'nursing care', the new calculatedCronbach's alpha coefficient would be up from 0.793 to 0.817. If move item B8 fromdimension 'know choice' into 'doctor services', the new calculated Cronbach's alpha coefficient was up to 0.748. So, as a conclusion, items in dimension 'admissionprocess' should be improvement.Conclusion: The OPSQ and IPSQ in this study were reliable and valid, it canbe use to evaluate the patient satisfaction in general hospital although it has still alarge room to improve.
|
PDF Full Text Request